Characterization

and Properties of
Petroleum Fractions
Characterization
and Properties of
Petroleum Fractions
First Edition

M. R. Riazi
Professor of Chemical Engineering
Kuwait University

P.O. Box 5969

Safat 13060, Kuwait
riazi@kuc01 .kuniv.edu.kw

ASTM Stock Number: MNL50

ASTM
100 Barr Harbor
West Conshohocken, PA 19428-2959

Printed in the U.S.A.

Library of Congress Cataloging-in-Publication Data

Riazi, M.-R.
Characterization and properties of petroleum fractions / M.-R. Riazi--1 st ed.
p. cm.--(ASTM manual series: MNL50)
ASTM stock number: MNL50
Includes bibliographical references and index.
ISBN 0-8031-3361-8
1. Characterization. 2. Physical property estimation. 3. Petroleum fractions--crude oils.

TP691.R64 2005
666.5---dc22
2004059586

Copyright 9 2005 AMERICAN SOCIETY FOR TESTING AND MATERIALS, West Conshohocken,
PA. All rights reserved. This material may not be reproduced or copied, in whole or in part, in any
printed, mechanical, electronic, film, or other distribution and storage media, without the written
consent of the publisher.

Photocopy Rights

Authorization to photocopy items for internal, personal, or educational classroom use,

or the internal, personal, or educational classroom use of specific clients, is granted by
the American Society for Testing and Materials (ASTM) provided that the appropriate fee
is paid to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923;
Tel: 508-750-8400; online: http://www.copyright.com/.

NOTE: This publication does not purport to address all of the safety problems associated with
its use. It is the responsibility of the user of this publication to establish appropriate safety and
health practices and determine the applicability of regulatory limitations prior to use.

Printed in Philadelphia, PA
January 2005
 To
Shiva, Touraj, and Nazly
Contents

Foreword xvii
Preface xix

Chapter 1--Introduction 1
Nomenclature 1
1.1 Nature of Petroleum Fluids 1
1.1.1 Hydrocarbons 3
1.1.2 Reservoir Fluids and Crude Oil 5
1.1.3 Petroleum Fractions and Products 7
1.2 Types and Importance of Physical Properties 10
1.3 Importance of Petroleum Fluids Characterization 12
1.4 Organization of the Book 15
1.5 Specific Features of this Manual 15
1.5.1 Introduction of Some Existing Books 15
1.5.2 Special Features of the Book 16
1.6 Applications of the Book 16
1.6.1 Applications in Petroleum Processing
(Downstream) 17
1.6.2 Applications in Petroleum Production
(Upstream) 17
1.6.3 Applications in Academia 17
1.6.4 Other Applications 17
1.7 Definition of Units and the Conversion Factors 17
1.7.1 Importance and Types of Units 17
1.7.2 Fundamental Units and Prefixes 18
1.7.3 Units of Mass 18
1.7.4 Units of Length 18
1.7.5 Units of Time 18
1.7.6 Units of Force 19
1.7.7 Units of Moles 19
1.7.8 Units of Molecular Weight 19
1.7.9 Units of Pressure 19
1.7.10 Units of Temperature 19
1.7.11 Units of Volume, Specific Volume, and
Molar Volume---The Standard Conditions 20
1.7.12 Units of Volumetric and Mass Flow Rates 20
1.7.13 Units of Density and Molar Density 20
1.7.14 Units of Specific Gravity 21
1.7.15 Units of Composition 21
1.7.16 Units of Energy and Specific Energy 22
1.7.17 Units of Specific Energy per Degrees 22
1.7.18 Units of Viscosity and Kinematic Viscosity 23
1.7.19 Units of Thermal Conductivity 23
1.7.20 Units of Diffusion Coefficients 23
1.7.21 Units of Surface Tension 24
1.7.22 Units of Solubility Parameter 24
1.7.23 Units of Gas-to-Oil Ratio 24
vii
viii CONTENTS

1.7.24 Values of Universal Constants 24

1.7.24.1 Gas Constant 24
1.7.24.2 Other Numerical Constants 24
1.7.25 Special Units for the Rates and Amounts of
Oil and Gas 24
1.8 Problems 26
References 27

Chapter 2--Characterization and Properties of Pure

Hydrocarbons 30
Nomenclature 30
2.1 Definition of Basic Properties 31
2.1.1 Molecular Weight 31
2.1.2 Boiling Point 31
2.1.3 Density, Specific Gravity, and API
Gravity 31
2.1.4 Refractive Index 32
2.1.5 Critical Constants (Tc, Pc, Vc, Zc) 32
2.1.6 Acentric Factor 33
2.1.7 Vapor Pressure 33
2.1.8 Kinematic Viscosity 33
2.1.9 Freezing and Melting Points 34
2.1.10 Flash Point 34
2.1.11 Autoignition Temperature 34
2.1.12 Flammability Range 34
2.1.13 Octane Number 34
2.1.14 Aniline Point 35
2.1.15 Watson K 35
2.1.16 Refractivity Intercept 35
2.1.17 Viscosity Gravity Constant 35
2.1.18 Carbon-to-Hydrogen Weight Ratio 36
2.2 Data on Basic Properties of Selected Pure
Hydrocarbons 36
2.2.1 Sources of Data 36
2.2.2 Properties of Selected Pure Compounds 37
2.2.3 Additional Data on Properties of Heavy
Hydrocarbons 37
2.3 Characterization of Hydrocarbons 45
2.3.1 Development of a Generalized Correlation
for Hydrocarbon Properties 45
2.3.2 Various Characterization Parameters for
Hydrocarbon Systems 48
2.3.3 Prediction of Properties of Heavy Pure
Hydrocarbons 50
2.3.4 Extension of Proposed Correlations to
Nonhydrocarbon Systems 54
2.4 Prediction of Molecular Weight, Boiling Point, and
Specific Gravity 55
2.4.1 Prediction of Molecular Weight 55
2.4.1.1 Riazi-Daubert Methods 55
2.4.1.2 ASTM Method 56
2.4.1.3 API Methods 56
2.4.1.4 Lee--Kesler Method 56
2.4.1.5 Goossens Correlation 57
2.4.1.6 Other Methods 58
 CONTENTS

2.4.2 Prediction of Normal Boiling Point 58

2.4.2.1 Riazi-Daubert Correlations 58
2.4.2.2 Soreide Correlation 58
2.4.3 Prediction of Specific Gravity/API Gravity 58
2.4.3.1 Riazi-Daubert Methods 58
2.5 Prediction of Critical Properties and Acentric
Factor 60
2.5.1 Prediction of Critical Temperature and
Pressure 60
2.5.1.1 Riazi-Daubert Methods 60
2.5.1.2 API Methods 60
2.5.1.3 Lee-Kesler Method 60
2.5.1.4 Cavett Method 61
2.5.1.5 Twu Method for To, Pc, Vc, and M 61
2.5.1.6 Winn-Mobil Method 62
2.5.1.7 Tsonopoulos Correlations 62
2.5.2 Prediction of Critical Volume 62
2.5.2.1 Riazi-Daubert Methods 62
2.5.2.2 Hall-Yarborough Method 63
2.5.2.3 API Method 63
2.5.3 Prediction of Critical Compressibility Factor 63
2.5.4 Prediction of Acentric Factor 64
2.5.4.1 Lee-Kesler Method 64
2.5.4.2 Edmister Method 65
2.5.4.3 Korsten Method 65
2.6 Prediction of Density, Refractive Index, CH Weight
Ratio, and Freezing Point 66
2.6.1 Prediction of Density at 20~C 66
2.6.2 Prediction of Refractive Index 66
2.6.3 Prediction of CH Weight Ratio 68
2.6.4 Prediction of Freezing/Melting Point 68
2.7 Prediction of Kinematic Viscosity at 38
and 99~ 70
2.8 The Winn Nomogram 73
2.9 Analysis and Comparison of Various
Characterization Methods 75
2.9.1 Criteria for Evaluation of a Characterization
Method 75
2.9.2 Evaluation of Methods of Estimation of
Molecular Weight 76
2.9.3 Evaluation of Methods of Estimation of
Critical Properties 77
2.9.4 Evaluation of Methods of Estimation of
Acentric Factor and Other Properties 81
2.10 Conclusions and Recommendations 82
2.11 Problems 83
References 84

Chapter 3--Characterization of Petroleum Fractions 87

Nomenclature 87
3.1 Experimental Data on Basic Properties of
Petroleum Fractions 88
3.1.1 Boiling Point and Distillation Curves 88
3.1.1.1 ASTM D86 88
3.1.1.2 True Boiling Point 89
x CONTENTS

3.1.1.3 Simulated Distillation by Gas

Chromatography 89
3.1.1.4 Equilibrium Flash Vaporization 91
3.1.1.5 Distillation at Reduced Pressures 92
3.1.2 Density, Specific Gravity, and API Gravity 93
3.1.3 Molecular Weight 93
3.1.4 Refractive Index 94
3.1.5 Compositional Analysis 95
3.1.5.1 Types of Composition 96
3.1.5.2 Analytical Instruments 96
3.1.5.3 PNA Analysis 98
3.1.5.4 Elemental Analysis 98
3.1.6 Viscosity 99
3.2 Prediction and Conversion of Distillation Data 100
3.2.1 Average Boiling Points 100
3.2.2 Interconversion of Various Distillation Data 101
3.2.2.1 Riazi-Daubert Method 102
3.2.2.2 Daubert's Method 103
3.2.2.3 Interconverion of Distillation Curves
at Reduced Pressures 106
3.2.2.4 Summary Chart for Interconverion
of Various Distillation Curves 108
3.2.3 Prediction of Complete Distillation Curves 108
3.3 Prediction of Properties of Petroleum Fractions 111
3.3.1 Matrix of Pseudocomponents Table 111
3.3.2 Narrow Versus Wide Boiling Range
Fractions 112
3.3.3 Use of Bulk Parameters (Undefined
Mixtures) 114
3.3.4 Method of Pseudocomponent (Defined
Mixtures) 114
3.3.5 Estimation of Molecular Weight, Critical
Properties, and Acentric Factor 115
3.3.6 Estimation of Density, Specific Gravity,
Refractive Index, and Kinematic Viscosity 116
3.4 General Procedure for Properties of Mixtures 119
3.4.1 Liquid Mixtures 119
3.4.2 Gas Mixtures 120
3.5 Prediction of the Composition of Petroleum
Fractions 120
3.5.1 Prediction of PNA Composition 120
3.5.1.1 Characterization Parameters for
Molecular Type Analysis 121
3.5.1.2 API Riazi-Daubert Methods 124
3.5.1.3 API Method 126
3.5.1.4 n-d-M Method 126
3.5.2 Prediction of Elemental Composition 127
3.5.2.1 Prediction of Carbon and Hydrogen
Contents 127
3.5.2.2 Prediction of Sulfur and Nitrogen
Contents 129
3.6 Prediction of Other Properties 130
3.6.1 Properties Related to Volatility 131
3.6.1.1 Reid Vapor Pressure 131
3.6.1.2 WL Ratio and Volatility Index 133
3.6.1.3 Flash Point 133
 CONTENTS

3.6.2 Pour Point 135

3.6.3 Cloud Point 135
3.6.4 Freezing Point 136
3.6.5 Aniline Point 137
3.6.5.1 Winn Method 137
3.6.5.2 Walsh-Mortimer 137
3.6.5.3 Linden Method 137
3.6.5.4 Albahri et al. Method 137
3.6.6 Cetane Number and Diesel Index 137
3.6.7 Octane Number 138
3.6.8 Carbon Residue 141
3.6.9 Smoke Point 142
3.7 Quality of Petroleum Products 143
3.8 Minimum Laboratory Data 143
3.9 Analysis of Laboratory Data and Development
of Predictive Methods 145
3.10 Conclusions and Recommendations 146
3.11 Problems 146
References 149

Chapter A Characterization o f Reservoir Fluids and

Crude Oils 152
Nomenclature 152
4.1 Specifications of Reservoir Fluids and Crude
Assays 153
4.1.1 Laboratory Data for Reservoir Fluids 153
4.1.2 Crude Oil Assays 154
4.2 Generalized Correlations for Pseudocritical
Properties of Natural Gases and Gas Condensate
Systems 160
4.3 Characterization and Properties of Single Carbon
Number Groups 161
4.4 Characterization Approaches for C7+ Fractions 163
4.5 Distribution functions for Properties of
Hydrocarbon-plus Fractions 164
4.5.1 General Characteristics 164
4.5.2 Exponential Model 165
4.5.3 Gamma Distribution Model 167
4.5.4 Generalized Distribution Model 170
4.5.4.1 Versatile Correlation 170
4.5.4.2 Probability Density Function for the
Proposed Generalized Distribution
Model 174
4.5.4.3 Calculation of Average Properties of
Hydrocarbon-Plus Fractions 175
4.5.4.4 Calculation of Average Properties of
Subfractions 177
4.5.4.5 Model Evaluations 178
4.5.4.6 Prediction of Property Distributions
Using Bulk Properties 181
4.6 Pseudoization and Lumping Approaches 184
4.6.1 Splitting Scheme 184
4.6.1.1 The Gaussian Quadrature Approach 185
4.6.1.2 Carbon Number Range Approach 186
4.6.2 Lumping Scheme 186
4.7 Continuous Mixture Characterization Approach 187
xii CONTENTS

4.8 Calculation of Properties of Crude Oils and

Reservoir Fluids 189
4.8.1 General Approach 190
4.8.2 Estimation of Sulfur Content of a Crude Oil 191
4.9 Conclusions and Recommendations 192
4.10 Problems 193
References 194

Chapter 5mPVT Relations and Equations of State 197

Nomenclature 197
5.1 Basic Definitions and the Phase Rule 198
5.2 PVT Relations 199
5.3 Intermolecular Forces 202
5.4 Equations of State 203
5.4.1 Ideal Gas Law 203
5.4.2 Real Gases--Liquids 203
5.5 Cubic Equations of State 204
5.5.1 Four Common Cubic Equations (vdW, RK,
SRK, and PR) 204
5.5.2 Solution of Cubic Equations of State 206
5.5.3 Volume Translation 207
5.5.4 Other Types of Cubic Equations of State 208
5.5.5 Application to Mixtures 209
5.6 Noncubic Equations of State 210
5.6.1 Virial Equation of State 210
5.6.2 Modified Benedict-Webb-Rubin Equation
of State 214
5.6.3 Carnahan-Starling Equation of State and Its
Modifications 214
5.7 Corresponding State Correlations 215
5.8 Generalized Correlation for PVT Properties of
Liquids--Rackett Equation 222
5.8.1 Rackett Equation for Pure Component
Saturated Liquids 222
5.8.2 Defined Liquid Mixtures and Petroleum
Fractions 223
5.8.3 Effect of Pressure on Liquid Density 223
5.9 Refractive Index Based Equation of State 225
5.10 Summary and Conclusions 227
5.11 Problems 228
References 229

Chapter 6---Thermodynamic Relations for Property Estimations 232

Nomenclature 232
6.1 Definitions and Fundamental Thermodynamic
Relations 234
6.1.1 Thermodynamic Properties and
Fundamental Relations 234
6.1.2 Measurable Properties 235
6.1.3 Residual Properties and Departure
Functions 236
6.1.4 Fugacity and Fugacity Coefficient for Pure
Components 237
6.1.5 General Approach for Property Estimation 238
6.2 Generalized Correlations for Calculation of
Thermodynamic Properties 238
 CONTENTS xiii

6.3 Properties of Ideal Gases 241

6.4 Thermodynamic Properties of Mixtures 247
6.4.1 Partial Molar Properties 248
6.4.2 Properties of Mixtures--Property Change
Due to Mixing 249
6.4.3 Volume of Petroleum Blends 251
6.5 Phase Equilibria of Pure Components--Concept
of Saturation Pressure 251
6.6 Phase Equilibria of Mixtures--Calculation
of Basic Properties 254
6.6.1 Definition of Fugacity, Fugacity Coefficient,
Activity, Activity Coefficient, and Chemical
Potential 254
6.6.2 Calculation of Fugacity Coefficients from
Equations of State 255
6.6.3 Calculation of Fugacity from Lewis Rule 256
6.6.4 Calculation of Fugacity of Pure Gases and
Liquids 256
6.6.5 Calculation of Activity Coefficients 257
6.6.6 Calculation of Fugacity of Solids 261
6.7 General Method for Calculation of Properties of
Real mixtures 263
6.8 Formulation of Phase Equilibria Problems for
Mixtures 263
6.8. I Criteria for Mixture Phase Equilibria 263
6.8.2 Vapor-Liquid Equilibria--Gas Solubility in
Liquids 265
6.8.2.1 Formulation of Vapor-Liquid
Equilibria Relations 265
6.8.2.2 Solubility of Gases in
Liquids--Henry's Law 266
6.8.2.3 Equilibrium Ratios (K/Values) 269
6.8.3 Solid-Liquid Equilibria--Solid Solubility 276
6.8.4 Freezing Point Depression and Boiling Point
Elevation 281
6.9 Use of Velocity of Sound in Prediction of Fluid
Properties 284
6.9.1 Velocity of Sound Based Equation
of State 286
6.9.2 Equation of State Parameters from Velocity
of Sound Data 287
6.9.2.1 Virial Coefficients 287
6.9.2.2 Lennard-Jones and van der Waals
Parameters 288
6.9.2.3 RK and PR EOS Parameters--
Property Estimation 289
6.10 Summary and Recommendations 292
6.11 Problems 292
References 294

Chapter 7--Applications: Estimation of Thermophysical

Properties 297
Nomenclature 297
7.1 General Approach for Prediction of
Thermophysical Properties of Petroleum Fractions
and Defined Hydrocarbon Mixtures 298
~v CONTENTS

7.2 Density 300

7.2.1 Density of Gases 300
7.2.2 Density of Liquids 300
7.2.3 Density of Solids 304
7.3 Vapor Pressure 305
7.3.1 Pure Components 305
7.3.2 Predictive Methods--Generalized
Correlations 306
7.3.3 Vapor Pressure of Petroleum Fractions 312
7.3.3.1 Analytical Methods 312
7.3.3.2 Graphical Methods for Vapor
Pressure of Petroleum Products
and Crude Oils 313
7.3.4 Vapor Pressure of Solids 314
7.4 Thermal Properties 316
7.4.1 Enthalpy 316
7.4.2 Heat Capacity 319
7.4.3 Heats of Phase Changes--Heat of
Vaporization 321
7.4.4 Heat of Combustion--Heating Value 324
7.5 Summary and Recommendations 326
7.6 Problems 327
References 328

Chapter 8mAppHcations: Estimation of Transport Properties 329

Nomenclature 329
8.1 Estimation of Viscosity 331
8.1.1 Viscosity of Gases 331
8.1.2 Viscosity of Liquids 335
8.2 Estimation of Thermal Conductivity 339
8.2.1 Thermal Conductivity of Gases 339
8.2.2 Thermal Conductivity of Liquids 342
8.3 Diffusion Coefficients 345
8.3.1 Diffusivity of Gases at Low Pressures 346
8.3.2 Diffusivity of Liquids at Low Pressures 347
8.3.3 Diffusivity of Gases and Liquids at High
Pressures 348
8.3.4 Diffusion Coefficients in Mutlicomponent
Systems 350
8.3.5 Diffusion Coefficient in Porous Media 350
8.4 Interrelationship Among Transport Properties 351
8.5 Measurement of Diffusion Coefficients in Reservoir
Fluids 354
8.6 Surface/Interracial Tension 356
8.6.1 Theory and Definition 356
8.6.2 Predictive Methods 358
8.7 Summary and Recommendations 361
8.8 Problems 362
References 362

Chapter 9--Applications: Phase Equilibrium Calculations 365

Nomenclature 365
9.1 Types of Phase Equilibrium Calculations 366
9.2 Vapor-Liquid Equilibrium Calculations 367
9.2.1 Flash Calculations--Gas-to-Oil Ratio 368
9.2.2 Bubble and Dew Points Calculations 370
 CONTENTS xv

9.2.3 Generation of P-T Diagrams--True Critical

Properties 372
9.3 Vapor-Liquid-Solid Equilibrium--Solid
Precipitation 373
9.3.1 Nature of Heavy Compounds, Mechanism of
their Precipitation, and Prevention Methods 373
9.3.2 Wax Precipitation--Solid Solution Model 378
9.3.3 Wax Precipitation: Multisolid-Phase
Model~Calculation of Cloud Point 382
9.4 Asphakene Precipitation: Solid-Liquid Equilibrium 385
9.5 Vapor-Solid Equilibrium--Hydrate Formation 388
9.6 Applications: Enhanced Oil Recovery--Evaluation
of Gas Injection Projects 390
9.7 Summary and Recommendations 391
9.8 Final Words 392
9.9 Problems 393
References 395
Appendix 397
Index 401
Foreword

THIS PUBLICATION,Characterization and Properties of Petroleum Fractions, was sponsored

by ASTM Committee D02 on Petroleum Fuels and Lubricants. The author is M. R. Riazi,
Professor of Chemical Engineering, Kuwait University, Safat, Kuwait. This publication
is Manual 50 of ASTM's manual series.

xvii
Preface

Scientists do not belong to any particular country, ideology, or religion, they belong
to the world community

THE FIELDOF Petroleum Characterization and Physical Properties has received significant
attention in recent decades with the expansion of computer simulators and advanced
analytical tools and the availability of more accurate experimental data. As a result of
globalization, structural changes are taking place in the chemical and petroleum indus-
try. Engineers working in these industries are involved with process simulators to design
and operate various units and equipment. Nowadays, a large number of process simula-
tors are being produced that are equipped with a variety of thermodynamic models and
choice of predictive methods for the physical properties. A person familiar with devel-
opment of such methods can make appropriate use of these simulators saving billions
of dollars in costs in investment, design, manufacture, and operation of various units
in these industries. Petroleum is a complex mixture of thousands of hydrocarbon com-
pounds and it is produced from an oil well in a form of reservoir fluid. A reservoir fluid is
converted to a crude oil through surface separation units and then the crude is sent to a
refinery to produce various petroleum fractions and hydrocarbon fuels such as kerosene,
gasoline, and fuel oil. Some of the refinery products are the feed to petrochemical plants.
More than half of world energy sources are from petroleum and probably hydrocarbons
will remain the most convenient and important source of energy and as a raw material
for the petrochemical plants at least throughout the 21 st century. Other fossil type fu-
els such as coal liquids are also mixtures of hydrocarbons although they differ in type
with petroleum oils. From 1970 to 2000, the share of Middle East in the world crude oil
reserves raised from 55 to 65%, but this share is expected to rise even further by 2010-
2020 when we near the point where half of oil reserves have been produced. The world
is not running out of oil yet but the era of cheap oil is perhaps near the end. Therefore,
economical use of the remaining oil and treatment of heavy oils become increasingly
important. As it is discussed in Chapter 1, use of more accurate physical properties for
petroleum fractions has a direct and significant impact on economical operation and
design of petroleum processing and production units which in turn would result in a
significant saving of existing petroleum reserves.
One of the most important tasks in petroleum refining and related processes is the
need for reliable values of the volumetric and thermodynamic properties for pure hy-
drocarbons and their mixtures. They are important in the design and operation of almost
every piece of processing equipment. Reservoir engineers analyze PVT and phase behav-
ior of reservoir fluids to estimate the amount of oil or gas in a reservoir, to determine an
optimum operating condition in a separator unit, or to develop a recovery process for
an oil or gas field. However, the most advanced design approaches or the most sophisti-
cated simulators cannot guarantee the optimum design or operation of a unit if required
input physical properties are not accurate. A process to experimentally determine the
volumetric, thermodynamic, and transport properties for all the industrially important
materials would be prohibitive in both cost and time; indeed it could probably never
be completed. For these reasons accurate estimations of these properties are becoming
increasingly important.
Characterization factors of m a n y types permeate the entire field of physical, ther-
modynamic, and transport property prediction. Average boiling points, specific gravity,
molecular weight, critical temperature, critical pressure, acentric factor, refractive index,
and certain molecular type analysis are basic parameters necessary to utilize methods
of correlation and prediction of the thermophysical properties. For correlating physi-
cal and thermodynamic properties, methods of characterizing undefined mixtures are

xix
xx PREFACE

necessary to provide input data. It could be imagined that the best method of character-
izing a mixture is a complete analysis. However, because of the complexity of undefined
mixtures, complete analyses are usually impossible and, at best, inconvenient. A predic-
tive method to determine the composition or amount of sulfur in a hydrocarbon fuel is
vital to see if a product meets specifications set by the government or other authorities
to protect the environment.
My first interaction with physical properties of petroleum fluids was at the time that
I was a graduate student at Penn State in the late 70s working on a project related to
enhanced oil recovery for my M.S. thesis when I was looking for methods of estimation
of properties of petroleum fluids. It was such a need and my personal interest that
later I joined the ongoing API project on thermodynamic and physical properties of
petroleum fractions to work for my doctoral thesis. Since that time, property estimation
and characterization of various petroleum fluids has remained one of my main areas of
research. Later in the mid-80s I rejoined Penn State as a faculty member and I continued
my work with the API which resulted in development of methods for several chapters
of the API Technical Data Book. Several years later in late 80s, I continued the work
while I was working at the Norwegian Institute of Technology (NTH) at Trondheim
where I developed some characterization techniques for heavy petroleum fractions as
well as measuring methods for some physical properties. In the 90s while at Kuwait
University I got the opportunity to be in direct contact with the oil companies in the
region through research, consultation, and conducting special courses for the industry.
My association with the University of Illinois at Chicago in early 90s was helpful in the
development of equations of state based on velocity of sound. The final revision of the
book was completed when I was associated with the University of Texas at Austin and
McGill University in Montreal during my leave from Kuwait University.
Characterization methods and estimating techniques presented in this book have been
published in various international journals or technical handbooks and included in
many commercial softwares and process simulators. They have also been presented as
seminars in different oil companies, universities, and research centers worldwide. The
major characteristics of these methods are simplicity, generality, accuracy, and avail-
ability of input parameters. Many of these methods have been developed by utilizing
some scientific fundamentals and combining them with a broad experimental data set
to generate semi-theoretical or semi-empirical correlations. Some of these methods have
been in use by the petroleum industry and research centers worldwide for the past two
decades.
Part of the materials in this book were prepared when I was teaching a graduate course
in applied thermodynamics in 1988 while at NTH. The materials, mainly a collection of
technical papers, have been continuously updated and rearranged to the present time.
These notes have also been used to conduct industrial courses as well as a course on fluid
properties in chemical and petroleum engineering. This book is an expansion with com-
plete revision and rewriting of these notes. The main objective of this book is to present
the fundamentals and practice of estimating the physical and thermodynamic proper-
ties as well as characterization methods for hydrocarbons, petroleum fractions, crude
oils, reservoir fluids, and natural gases, as well as coal liquids. However, the emphasis is
on the liquid petroleum fractions, as properties of gases are generally calculated more
accurately. The book will emphasize manual calculations with practical problems and
examples but also wilI provide good understanding of techniques used in commercial
software packages for property estimations. Various methods and correlations developed
by different researchers commonly used in the literature are presented with necessary
discussions and recommendations.
My original goal and objective in writing this book was to provide a reference for the
petroleum industry in both processing and production. It is everyone's experience that in
using thermodynamic simulators for process design and equipment, a large number of
options is provided to the user for selection of a method to characterize the oil or to get
an estimate of a physical property. This is a difficult choice for a user of a simulator, as
the results of design calculations significantly rely on the method chosen to estimate the
properties. One of my goals in writing this book was to help users of simulators overcome
this burden. However, the book is written in a way that it can also be used as a textbook
for graduate or senior undergraduate students in chemical, petroleum, or mechanical
engineering to understand the significance of characterization, property estimation and
 PREFACE xxi

methods of their development. For this purpose a set of problems is presented at the
end of each chapter. The book covers characterization as well as methods of estimation
of thermodynamic and transport properties of various petroleum fluids and products. A
great emphasis is given to treatment of heavy fractions throughout the book. An effort
was made to write the book in a way that not only would be useful for the profession-
als in the field, but would also be easily understandable to those non-engineers such as
chemists, physicists, or mathematicians who get involved with the petroleum industry.
The word properties in the title refers to thermodynamic, physical, and transport proper-
ties. Properties related to the quality and safety of petroleum products are also discussed.
Organization of the book, its uses, and importance of the methods are discussed in detail
in Chapter 1. Introduction of similar books and the need for the present book as well as
its application in the industry and academia are also discussed in Chapter 1. Each chap-
ter begins with nomenclature and ends with the references used in that chapter. Exercise
problems in each chapter contain additional information and methods. More specific
information about each chapter and its contents are given in Chapter 1. As Goethe
said, "Things which matter most must never be at the mercy of things which matter
least."
I am indebted to many people especially teachers, colleagues, friends, students, and,
above all, my parents, who have been so helpful throughout my academic life. I am partic-
ularly thankful to Thomas E. Daubert of Pennsylvania State University who introduced
to me the field of physical properties and petroleum characterization in a very clear and
understandable way. Likewise, I am thankful to Farhang Shadman of the University of
Arizona who for the first time introduced me to the field of chemical engineering re-
search during my undergraduate studies. These two individuals have been so influential
in shaping my academic life and I am so indebted to them for their h u m a n characters
and their scientific skills. I have been fortunate to meet and talk with many scientists and
researchers from both the oil industry and academia from around the world during the
last two decades whose thoughts and ideas have in m a n y ways been helpful in shaping
the book.
I am also grateful to the institutions, research centers, and oil companies that I have
been associated with or that have invited me for lecturing and consultation. Thanks to
Kuwait University as well as Kuwait Petroleum Corporation (KPC) and KNPC, m a n y of
whose engineers I developed working relations with and have been helpful in evaluation
of many of the estimating methods throughout the years. I am thankful to all scientists
and researchers whose works have been used in this book and I hope that all have
been correctly and appropriately cited. I would be happy to receive their comments and
suggestions on the book. Financial support from organizations such as API, NSF, GPA,
GRI, SINTEE Petrofina Exploration Norway, NSERC Canada, Kuwait University, and
KFAS that was used for my research work over the past two decades is also appreciated.
I am grateful to ASTM for publishing this work and particularly to Geroge Totten who
was the first to encourage me to begin writing this book. His advice, interest, support,
and suggestions through the long years of writing the book have been extremely helpful
in completing this project. The introductory comments from him as well as those from
Philip T. Eubank and Jos6 Luis Pefia Diez for the back cover are appreciated. I a m also
grateful to the four unanimous reviewers who tirelessly reviewed the entire and lengthy
manuscript with their constructive comments and suggestions which have been reflected
in the book. Thanks also to Kathy Dernoga, the publishing manager at ASTM, who was
always cooperative and ready to answer my questions and provided me with necessary
information and tools during the preparation of this manuscript. Her encouragements
and assistance were quite useful in pursuing this work. She also was helpful in the de-
sign of the front and back covers of the book as well as providing editorial suggestions.
I am thankful to Roberta Storer and Joe Ermigiotti for their excellent job of editing and
updating the manuscript. Cooperation of other ASTM staff, especially Monica Siperko,
Carla J. Falco, and Marsha Firman is highly appreciated. The art work and most of
the graphs and figures were prepared by Khaled Damyar of Kuwait University and his
efforts for the book are appreciated. I also sincerely appreciate the publishers and the
organizations that gave their permissions to include some published materials, in partic-
ular API, ACS, AIChE, GPA, Elsevier (U.K.), editor of Oil & Gas J., McGraw-Hill, Marcel
and Dekker, Wiley, SPE, and Taylor and Francis. Thanks to the manager and personnel
of KISR for allowing the use of photos of their instruments in the book. Finally and
xxii PREFACE

most importantly, I must express my appreciation and thanks to my family who have
been helpful and patient during all these years and without whose cooperation, moral
support, understanding, and encouragement this project could never have been under-
taken. This book is dedicated to my family, parents, teachers, and the world scientific
community.

M. R. Riazi
August 2004
 MNL50-EB/Jan. 2005

1
Introduction

NOMENCLATURE Greek Letters

Relative volatility
API API gravity
~0 Fugacity coefficient
A% Percent of aromatics in a petroleum
a~ Acentric factor
fraction
Surface tension
D Diffusion coefficient
p Density at temperature T and pressure P
CH Carbon-to-hydrogen weight ratio
/~ Viscosity
d Liquid density at 20~ and 1 atm
v Kinematic viscosity
Kw Watson K factor
k Thermal conductivity
Ki Equilibrium ratio of component i in Acronyms
a mixture API-TDB American Petroleum Institute-Technical Data
log10 Logarithm of base l0
Book
In Logarithm of base e
bbl Barrel
M Molecular weight
GOR Gas-to-oil ratio
Nmin Minimum number of theoretical plates IUPAC International Union of Pure and Applied Chem-
in a distillation column
istry
N% Percent of naphthenes in a petroleum
PNA Paraffin, naphthene, aromatic content of a
fraction
petroleum fraction
n Sodium D line refractive index of liquid
SC Standard conditions
at 20~ and 1 atrn, dimensionless
scf Standard cubic feet
n Number of moles
stb Stock tank barrel
P Pressure
STO Stock tank oil
Pc Critical pressure
STP Standard temperature and pressure
psat Vapor (saturation) pressure
P% Percent of paraffins in a petroleum the nature of petroleum
IN THIS INTRODUCTORY CHAPTER, f i r s t
fraction fluids, hydrocarbon types, reservoir fluids, crude oils, natural
R Universal gas constant gases, and petroleum fractions are introduced and then types
Ri Refractivity intercept and importance of characterization and physical properties
SG Specific gravity at 15.5~ (60~ are discussed. Application of materials covered in the book in
SUS Saybolt Universal Seconds (unit of various parts of the petroleum industry or academia as well
viscosity) as organization of the book are then reviewed followed by
S% Weight % of sulfur in a petroleum specific features of the book and introduction of some other
fraction related books. Finally, units and the conversion factors for
T Temperature those parameters used in this book are given at the end of the
Tb Boiling point chapter.
Tc Critical temperature
TF Flash point
Tp Pour point 1.1 NATURE OF PETROLEUM FLUIDS
TM Melting (freezing point) point
V Volume Petroleum is one of the most important substances consumed
Xm Mole fraction of a component in by m a n at present time. It is used as a main source of energy
a mixture for industry, heating, and transportation and it also pro-
Xv Volume fraction of a component in vides the raw materials for the petrochemical plants to pro-
a mixture duce polymers, plastics, and m a n y other products. The word
Xw Weight fraction of a component in a petroleum, derived from the Latin words petra and oleum,
mixture means literally rock oil and a special type of oil called oleum
y Mole fraction of a component in a vapor [1]. Petroleum is a complex mixture of hydrocarbons that
phase occur in the sedimentary rocks in the form of gases (natural

Copyright 9 2005 by ASTM International www.astm.org

2 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
gas), liquids (crude oil), semisolids (bitumen), or solids (wax
or asphaltite). Liquid fuels are normally produced from liq-
uid hydrocarbons, although conversion of nonliquid hydro-
carbons such as coal, oil shale, and natural gas to liquid fuels
is being investigated. In this book, only petroleum hydrocar-
bons in the form of gas or liquid, simply called petroleum flu-
ids, are considered. Liquid petroleum is also simply called oil.
Hydrocarbon gases in a reservoir are called a natural gas or
simply a gas. An underground reservoir that contains hydro-
carbons is called petroleum reservoir and its hydrocarbon con-
tents that can be recovered through a producing well is called
reservoir fluid. Reservoir fluids in the reservoirs are usually in
contact with water in porous media conditions and because
they are lighter than water, they stay above the water level
under natural conditions.
Although petroleum has been known for many centuries,
the first oil-producing well was discovered in 1859 by E.L.
Drake in the state of Pennsylvania and that marked the
birth of modern petroleum technology and refining. The
main elements of petroleum are carbon (C) and hydrogen
(H) and some small quantities of sulfur (S), nitrogen (N),
and oxygen (O). There are several theories on the formation
of petroleum. It is generally believed that petroleum is de-
rived from aquatic plants and animals through conversion of
organic compounds into hydrocarbons. These animals and
plants under aquatic conditions have converted inorganic
compounds dissolved in water (such as carbon dioxide) to
organic compounds through the energy provided by the sun.
An example of such reactions is shown below:
(1.1) 6CO2 + 6H20 d- energy --~ 602 + C6H1206
in which C6H1206 is an organic compound called carbohy-
drate. In some cases organic compounds exist in an aquatic
environment. For example, the Nile river in Egypt and the
Uruguay river contain considerable amounts of organic ma-
terials. This might be the reason that most oil reservoirs are
located near the sea. The organic compounds formed may be
decomposed into hydrocarbons under certain conditions.
(1.2) (CHEO)n --~ xCO2 d-yCH4
in which n, x, y, and z are integer numbers and yCHz is the
closed formula for the produced hydrocarbon compound.
Another theory suggests that the inorganic compound cal-
cium carbonate (CaCO3) with alkali metal can be converted to
calcium carbide (CaC2), and then calcium carbide with water
(H20) can be converted to acetylene (C2H2). Finally, acetylene
can be converted to petroleum [ 1]. Conversion of organic mat-
ters into petroleum is called maturation. The most important
factors in the conversion of organic compounds to petroleum
hydrocarbons are (1) heat and pressure, (2) radioactive rays,
such as g a m m a rays, and (3) catalytic reactions. Vanadium-
and nickel-type catalysts are the most effective catalysts in
the formation of petroleum. For this reason some of these
metals may be found in small quantities in petroleum fluids.
The role of radioactive materials in the formation of hydro-
carbons can be best observed through radioactive bombard-
ing of fatty acids (RCOOH) that form paraffin hydrocarbons.
Occasionally traces of radioactive materials such as uranium
and potassium can also be found in petroleum. In summary,
the following steps are required for the formation of hydrocar-
bons: (1) a source of organic material, (2) a process to convert
organic compounds into petroleum, and (3) a sealed reservoir
space to store the hydrocarbons produced. The conditions re-
quired for the process of conversion of organic compounds
into petroleum (as shown through Eq. (1.2) are (1) geologic
time of about 1 million years, (2) m a x i m u m pressure of
about 17 MPa (2500 psi), and (3) temperature not exceed-
ing 100-120~ (~210-250~ If a leak occurred sometime
in the past, the exploration well will encounter only small
amounts of residual hydrocarbons. In some cases bacteria
may have biodegraded the oil, destroying light hydrocarbons.
An example of such a case would be the large heavy oil accu-
mulations in Venezuela. The hydrocarbons generated grad-
ually migrate from the original beds to more porous rocks,
such as sandstone, and form a petroleum reservoir. A series
of reservoirs within a c o m m o n rock is called an oil field.
Petroleum is a mixture of hundreds of different identifiable
hydrocarbons, which are discussed in the next section. Once
petroleum is accumulated in a reservoir or in various sedi-
ments, hydrocarbon compounds may be converted from one
form to another with time and varying geological conditions.
This process is called in-situ alteration, and examples of chem-
ical alteration are thermal maturation and microbial degra-
dation of the reservoir oil. Examples of physical alteration of
petroleum are the preferential loss of low-boiling constituents
by the diffusion or addition of new materials to the oil in
place from a source outside the reservoir [1]. The main dif-
ference between various oils from different fields around the
world is the difference in their composition of hydrocarbon
compounds. Two oils with exactly the same composition have
identical physical properties under the same conditions [2].
A good review of statistical data on the amount of oil and
gas reservoirs, their production, processing, and consump-
tion is usually reported yearly by the Oil and Gas Journal
(OGJ). An annual refinery survey by OGJ is usually published
in December of each year. OGJ also publishes a forecast and
review report in January and a midyear forecast report in
July of each year. In 2000 it was reported that total proven oil
reserves is estimated at 1016 billion bbl (1.016 x 10 tz bbl),
which for a typical oil is equivalent to approximately 1.39 x
1011 tons. The rate of oil production was about 64.6 million
bbl/d (~3.23 billion ton/year) through more than 900 000 pro-
ducing wells and some 750 refineries [3, 4]. These numbers
vary from one source to another. For example, Energy Infor-
mation Administration of US Department of Energy reports
world oil reserves as of January 1, 2003 as 1213.112 billion
bbl according to OGJ and 1034.673 billion bbl according to
World Oil (www.eia.doe.gov/emeu/iea). According to the OGJ
worldwide production reports (Oil and Gas Journal, Dec. 22,
2003, p. 44), world oil reserves estimates changed from 999.78
in 1995 to 1265.811 billion bbl on January 1, 2004. For the
same period world gas reserves estimates changed from 4.98 x
1015 scf to 6.0683 x 1015 scf. In 2003 oil consumption was
about 75 billion bbl/day, and it is expected that it will in-
crease to more than 110 million bbl/day by the year 2020.
This means that with existing production rates and reserves,
it will take nearly 40 years for the world's oil to end. Oil
reserves life (reserves-to-production ratio) in some selected
countries is given by OGJ (Dec. 22, 2004, p. 45). According
to 2003 production rates, reserves life is 6.1 years in UK,
10.9 years in US, 20 years in Russia, 5.5 years in Canada,
84 years in Saudi Arabia, 143 years in Kuwait, and 247 years
 1. INTRODUCTION 3

in Iraq. As in January l, 2002, the total number of world oil interconversion schemes may occur during the formation,
wells was 830 689, excluding shut or service wells (OGJ, Dec. maturation, and in-situ alteration of petroleum.
22, 2004). Estimates of world oil reserves in 1967 were at Paraffins are also called alkanes and have the general for-
418 billion and in 1987 were at 896 billion bbl, which shows mula of C, Han+a, where n is the number of carbon atoms.
an increase of 114% in this period [5]. Two-thirds of these Paraffins are divided into two groups of normal and isoparaf-
reserves are in the Middle East, although this portion de- fins. Normal paraffins or normal alkanes are simply written
pends on the type of oil considered. Although some people as n-paraffins or n-alkanes and they are open, straight-chain
believe the Middle East has a little more than half of world saturated hydrocarbons. Paraffins are the largest series of hy-
oil reserves, it is believed that many undiscovered oil reser- drocarbons and begin with methane (CH4), which is also rep-
voirs exist offshore under the sea, and with increase in use resented by C1. Three n-alkanes, methane (C1), ethane (C2),
of the other sources of energy, such as natural gas or coal, and n-butane (C4), are shown below:
and through energy conservation, oil production may well
continue to the end of the century. January 2000, the total H H H H H H H
amount of gas reserves was about 5.15 • 1015 scf, and I I I I I I I
H--C--H H--C--C--H H--C--C--C--C--H
its production in 1999 was about 200 x 109 scf/d (5.66 x
109 sm3/d) through some 1500 gas plants [3]. In January I I I I I I I
H H H H H H H
2004, according to OGJ (Dec. 22, 2004, p. 44), world natu-
ral gas reserves stood at 6.068 • 1015 scf (6068.302 trillion
scf). This shows that existing gas reserves may last for some Methane Ethane n-Butane
70 years. Estimated natural gas reserves in 1975 were at
2.5 x 1015 scf (7.08 x 1013 sm3), that is, about 50% of current (CH4) (C2H6) (C4H1~
reserves [6]. In the United States, consumption of oil and gas
The open formula for n-C4 can also be shown as CH3--
in 1998 was about 65% of total energy consumption. Crude
CH2--CH2--CH3 and for simplicity in drawing, usually the
oil demand in the United State in 1998 was about 15 million
CH3 and CH2 groups are not written and only the carbon-
bbl/d, that is, about 23% of total world crude production [3].
Worldwide consumption of natural gas as a clean fuel is on carbon bonds are drawn. For example, for a n-alkane com-
pound of n-heptadecane with the formula of C17H36, the
the rise, and attempts are underway to expand the trans-
structure can also be shown as follows:
fer of natural gas through pipelines as well as its conver-
sion to liquid fuels such as gasoline. The world energy con-
sumption is distributed as 35% through oil, 31% through
coal, and 23% through natural gas. Nearly 11% of total
world energy is produced through nuclear and hydroelectric n-Heptadecane (C17H36)
sources [ 1].
The second group of paraffins is called isoparaffins; these
are branched-type hydrocarbons and begin with isobutane
(methylpropane), which has the same closed formula as n-
1.1.1 Hydrocarbons butane (Call10). Compounds of different structures but the
In early days of chemistry science, chemical compounds were same closed formula are called isomers. Three branched or
divided into two groups: inorganic and organic, depending isoparaffin compounds are shown below:
on their original source. Inorganic compounds were obtained
from minerals, while organic compounds were obtained from CH3 CH3 CH3
living organisms and contained carbon. However, now or- CH3--CH--CH3 CH3--CH~CH2--CH3 CH3--CH--CH2--CH2--CH2--CH2--CH3
ganic compounds can be produced in the laboratory. Those
organic compounds that contain only elements of carbon (C) isobutane isopen~ane
(methylbutane) isooctane(2-methylheptane)
and hydrogen (H) are called hydrocarbons, and they form
the largest group of organic compounds. There might be as (C4HIo) (C5H12) (C8HI8)
many as several thousand different hydrocarbon compounds
in petroleum reservoir fluids. Hydrocarbon compounds have In the case of isooctane, if the methyl group (CH3) is at-
a general closed formula of CxHy, where x and y are integer tached to another carbon, then we have another compound
numbers. The lightest hydrocarbon is methane (CH4), which (i.e., 3-methylheptane). It is also possible to have more than
is the main component in a natural gas. Methane is from a one branch of CH3 group, for example, 2,3-dimethylhexane
group of hydrocarbons called paraffins. Generally, hydrocar- and 2-methylheptane, which are simply shown as following:
bons are divided into four groups: (1) paraffins, (2) olefins,
(3) naphthenes, and (4) aromatics. Paraffins, olefins, and
naphthenes are sometime called aliphatic versus aromatic
compounds. The International Union of Pure and Applied
Chemistry (IUPAC) is a nongovernment organization that
provides standard names, nomenclature, and symbols for dif- 2-Methylheptane (CsHls) 2,3-Dimethylhexane (C8H18)
ferent chemical compounds that are widely used [7]. The
relationship between the various hydrocarbon constituents Numbers refer to carbon numbers where the methyl group
of crude oils is hydrogen addition or hydrogen loss. Such is attached. For example, 1 refers to the first carbon either
4 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

1.0E+15 Unsaturated compounds are more reactive than saturated hy-

drocarbons (without double bond). Olefins are uncommon in
crude oils due to their reactivity with hydrogen that makes
them saturated; however, they can be produced in refiner-
ies through cracking reactions. Olefins are valuable prod-
I.OE+IO ucts of refineries and are used as the feed for petrochemical
plants to produce polymers such as polyethylene. Similarly
compounds with triple bonds such as acetylene (CH------CH)are
not found in crude oils because of their tendency to become
saturated [2].
N a p h t h e n e s or cycloalkanes are ring or cyclic saturated hy-
1.0E~5
drocarbons with the general formula of CnH2n. Cyclopentane
(C5H10), cyclohexane (C6H12), and their derivatives such as
n-alkylcyclopentanes are normally found in crude oils. Three
types of naphthenic compounds are shown below:

1.0E+O0
0 10 20 30 40 50
Number of CarbonAtoms
FIG. 1.1reNumber of possible alkane isomers.
Cyclopentane Methylcyclopentane Ethylcyclohexane
from the right or from the left. There are 2 isomers for bu- (CsHIo) (C6HI2) (C8H~6)
tane and 3 for pentane, but there are 5 isomers for hexane, 9
for heptane, 18 for octane (C8H18), and 35 for nonane. Sim-
If there is only one alkyl group from n-paraffins (i.e., methyl,
ilarly, dodecane (C12H26) has 355, while octadecane (C18H38)
ethyl, propyl, n-butyl .... ) attached to a cyclopentane hydro-
has 60523 and C40 has 62 x 1012 isomers [1, 8, 9]. The num-
carbon, the series is called n-alkylcyclopentanes, such as the
ber of isomers rapidly increases with the number of carbon
two hydrocarbons shown above where on each junction of the
atoms in a molecule because of the rapidly rising number of
ring there is a CH2 group except on the alkyl group juncture
their possible structural arrangements as shown in Fig. 1.1.
where there is only a CH group. For simplicity in drawing,
For the paraffins in the range of Cs-C12, the number of iso-
these groups are not shown. Similarly there is a homologous
mers is more than 600 although only about 200-400 of them
have been identified in petroleum mixtures [ 10]. Isomers have napthenic series of n-alkylcyclohexanes with only one satu-
different physical properties. The same increase in number rated ring of cyclohexane, such as ethylcyclohexane shown
of isomers with molecular weight applies to other hydro- above. Napthenic hydrocarbons with only one ring are also
carbon series. As an example, the total number of hydrocar- called monocycloparaffins or mononaphthenes. In heavier oils,
bons (from different groups) having 20 carbon atoms is more saturated multirings attached to each other called polycy-
than 300000 [10]! cloparaffins orpolynaphthenes may also be available. Thermo-
Under standard conditions (SC) of 20~ and 1 atm, the dynamic studies show that naphthene rings with five and six
carbon atoms are the most stable naphthenic hydrocarbons.
first four members of the alkane series (methane, ethane,
The content of cycloparaffins in petroleum may vary up to
propane, and butane) are in gaseous form, while from C5Hl1
(pentane) to n-hexadecane (C16H36)they are liquids, and from 60%. Generally, any petroleum mixture that has hydrocarbon
n-heptadecane (C17H38) the compounds exist as waxlike solids compounds with five carbon atoms also contains naphthenic
at this standard temperature and pressure. Paraffins from C1 compounds.
A r o m a t i c s are an important series of hydrocarbons found
to C40 usually appear in crude oil and represent up to 20% of
crude by volume. Since paraffins are fully saturated (no dou- in almost every petroleum mixture from any part of the world.
ble bond), they are stable and remain unchanged over long Aromatics are cyclic but unsaturated hydrocarbons that begin
periods of geological time. with benzene molecule (C6H6) and contain carbon-carbon
Olefms are another series of noncyclic hydrocarbons but double bonds. The name aromatic refers to the fact that such
they are unsaturated and have at least one double bond hydrocarbons commonly have fragrant odors. Four different
between carbon-carbon atoms. Compounds with one dou- aromatic compounds are shown below:
ble bond are called monoolefins or alkenes, such as ethene
\
(also named ethylene: CH2=CH2) and propene or propylene
(CH2=CH--CH3). Besides structural isomerism connected
with the location of double bond, there is another type of iso-
merism called geometric isomerism, which indicates the way
9
atoms are oriented in space. The configurations are differen-
tiated in their names by the prefixes cis- and trans- such as (C6H6) (C7H8) (C8H1o) (C1o~8)
cis- and trans-2-butene. Monoolefins have a general formula
Benzene Toluene O-xylene Naphthalene
of CnH2n. If there are two double bonds, the olefin is called
diolefin (or diene), such as butadiene (CH2=CH--CH=CH2). (Methylbenzene) ( 1,2-Dimethylbenzene)

In the above structures, on each junction on the benzene
ring where there are three bonds, there is only a group of CH,
while at the junction with an alkylgroup (i.e., toluene) there
is only a C atom. Although benzene has three carbon-carbon
double bonds, it has a unique arrangement of electrons that
allows benzene to be relatively unreactive. Benzene is, how-
ever, known to be a cancer-inducing compound [2]. For this
reason, the amount of benzene allowed in petroleum prod-
ucts such as gasoline or fuel oil is limited by government
regulations in many countries. Under SC, benzene, toluene,
and xylene are in liquid form while naphthalene is in a solid
state. Some of the common aromatics found in petroleum
and crude oils are benzene and its derivatives with attached
methyl, ethyl, propyl, or higher alkyl groups. This series of
aromatics is called alkylbenzenes and compounds in this ho-
mologous group of hydrocarbons have a general formula
of CnH2n-6 (where n _> 6). Generally, aromatic series with
only one benzene ring are also called monoaromatics (MA)
or mononuclear aromatics. Naphthalene and its derivatives,
which have only two unsaturated rings, are sometime called
diaromatics. Crude oils and reservoir fluids all contain aro-
matic compounds. However, heavy petroleum fractions and
residues contain multi-unsaturated rings with many benzene
and naphthene rings attached to each other. Such aromatics
(which under SC are in solid form) are also calledpolyaromat-
ics (PA) or polynuclear aromatics (PNA). In this book terms of
mono and polyaromatics are used. Usually, heavy crude oils
contain more aromatics than do light crudes. The amount of
aromatics in coal liquids is usually high and could reach as
high as 98% by volume. It is common to have compounds
with napthenic and aromatic rings side by side, especially
in heavy fractions. Monoaromatics with one napthenic ring
have the formula of CnH2n-8 and with two naphthenic rings
the formula is C~Hzn-8. There are many combinations of alkyl-
naphthenoaromatics [ 1, 7].
Normally, high-molecular-weight polyaromatics contain
several heteroatoms such as sulfur (S), nitrogen (N), or oxygen
(O) hut the compound is still called an aromatic hydrocarbon.
Two types of these compounds are shown below [1 ]:
H
Dibenzothiophene Benzocarbazole (CI6H1IN)
Except for the atoms S and N, which are specified in the above
structures, on other junctions on each ring there is either a
CH group or a carbon atom. Such heteroatoms in multiring
aromatics are commonly found in asphaltene compounds as
shown in Fig. 1.2, where for simplicity, C and H atoms are not
shown on the rings.
Sulfur is the most important heteroatom in petroleum and
it can be found in cyclic as well as noncyclic compounds such
as mercaptanes (R--S--H) and sulfides (R--S--W), where R
and R' are alkyl groups. Sulfur in natural gas is usually found
in the form of hydrogen sulfide (H2S). Some natural gases
1. INTRODUCTION 5
C: 8 3 . 1 %
H: 8.9%
N: 1.0%
O: 0%
S: 7 . 0 %
H/C: i.28
Molecular Weighh 1370
FIG. 1.2mAn example of asphaltene molecule. Reprinted from
Ref. [1], p. 463, by courtesy of Marcel Dekker, Inc.
contain HzS as high as 30% by volume. The amount of sulfur
in a crude may vary from 0.05 to 6% by weight. In Chapter 3,
further discussion on the sulfur contents of petroleum frac-
tions and crude oils will be presented. The presence of sulfur
in finished petroleum products is harmful, for example, the
presence of sulfur in gasoline can promote corrosion of en-
gine parts. Amounts of nitrogen and oxygen in crude oils are
usually less than the amount of sulfur by weight. In general
for petroleum oils, it appears that the compositions of ele-
ments vary within fairly narrow limits; on a weight basis they
are [1]
Carbon (C), 83.0-87.0%
Hydrogen (H), 10.0-14.0%
Nitrogen (N), 0.1-2.0%
Oxygen (O), 0.05-1.5%
Sulfur (S), 0.05-6.0%
Metals (Nickel, Vanadium, and Copper), < 1000 ppm (0.1%)
Generally, in heavier oils (lower API gravity, defined by
Eq. (2.4)) proportions of carbon, sulfur, nitrogen, and oxygen
elements increase but the amount of hydrogen and the overall
quality decrease. Further information and discussion about
the chemistry of petroleum and the type of compounds found
in petroleum fractions are given by Speight [ 1]. Physical prop-
erties of some selected pure hydrocarbons from different ho-
mologous groups commonly found in petroleum fluids are
given in Chapter 2. Vanadium concentrations of above 2 ppm
in fuel oils can lead to severe corrosion in turbine blades and
deterioration of refractory in furnaces. Ni, Va, and Cu can also
severely affect the activities of catalysts and result in lower
products. The metallic content may be reduced by solvent
extraction with organic solvents. Organometallic compounds
are precipitated with the asphaltenes and residues.
1.1.2 Reservoir Fluids and Crude Oil
The word fluid refers to a pure substance or a mixture of com-
pounds that are in the form of gas, liquid, or both a mixture
of liquid and gas (vapor). Reservoir fluid is a term used for the
mixture of hydrocarbons in the reservoir or the stream leaving
a producing well. Three factors determine if a reservoir fluid is
in the form of gas, liquid, or a mixture of gas and liquid. These
factors are (1) composition of reservoir fluid, (2) temperature,
and (3) pressure. The most important characteristic of a reser-
voir fluid in addition to specific gravity (or API gravity) is its
gas-to-oil ratio (GOR), which represents the amount of gas
6 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

TABLE 1.1--Types and characteristics of various reservoir fluids.

Reservoirfluidtype GOR, scf/sth CH4, mol% C6+,tool% API gravityof STOa
Black oil <1000 _<50 >_30 <40
Volatile oil 1000-3000 50-70 10-30 40-45
Gas condensate 3000-50 000 70-85 3-10 _>45
Wet gas _>50000 >-75 <3 >50
Dry gas >-10 0000 >_90 <1 No liquid
"API gravityof stock tank oil (STO)produced at the surface facilitiesat standard conditions(289 K and 1 atm).

p r o d u c e d at SC i n s t a n d a r d cubic feet (scf) to the a m o u n t of
liquid oil produced at the SC i n stock t a n k barrel (stb). Other
u n i t s of GOR are discussed i n Section 1.7.23 a n d its calcula-
tion is discussed i n Chapter 9. Generally, reservoir fluids are
categorized into four or five types (their characteristics are
given in Table 1.1). These five fluids i n the direction of in-
creasing GOR are black oil, volatile oil, gas condensate, wet
gas, and dry gas.
If a gas after surface separator, u n d e r SC, does n o t pro-
duce any liquid oil, it is called dry gas. A n a t u r a l gas that after
p r o d u c t i o n at the surface facilities can produce a little liquid
oil is called wet gas. The word wet does n o t m e a n that the
gas is wet with water, b u t refers to the h y d r o c a r b o n liquids
that condense at surface conditions. I n dry gases n o liquid
h y d r o c a r b o n is formed at the surface conditions. However,
both dry a n d wet gases are i n the category of n a t u r a l gases.
Volatile oils have also been called high-shrinkage crude oil a n d
near-critical oils, since the reservoir t e m p e r a t u r e a n d pressure
are very close to the critical p o i n t of such oils, b u t the critical
t e m p e r a t u r e is always greater t h a n the reservoir t e m p e r a t u r e
[i 1]. Gases a n d gas c o n d e n s a t e fluids have critical tempera-
tures less t h a n the reservoir temperature. Black oils c o n t a i n
heavier c o m p o u n d s a n d therefore the API gravity of stock
t a n k oil is generally lower t h a n 40 a n d the GOR is less t h a n
1000 scf/stb. The specifications given i n Table 1.1 for various
reservoir fluids, especially at the b o u n d a r i e s b e t w e e n differ-
ent types, are arbitrary a n d vary from one source to a n o t h e r
[9, 11]. It is possible to have a reservoir fluid type that has
properties outside the corresponding limits m e n t i o n e d ear-
lier. D e t e r m i n a t i o n of a type of reservoir fluid by the above
rule of t h u m b based on the GOR, API gravity of stock t a n k
oil, or its color is not possible for all fluids. A more accu-
rate m e t h o d of d e t e r m i n i n g the type of a reservoir fluid is
based o n the phase behavior calculations, its critical point,
a n d shape of the phase d i a g r a m which will be discussed i n
Chapters 5 a n d 9. I n general, oils produced from wet gas,
gas condensate, volatile oil, a n d black oil increase i n spe-
cific gravity (decrease in API gravity a n d quality) in the same
order. Here quality of oil indicates lower carbon, sulfur, nitro-
gen, a n d metal contents which correspond to higher heating
value. Liquids from black oils are viscous a n d black i n color,
while the liquids from gas condensates or wet gases are clear
a n d colorless. Volatile oils produce fluids b r o w n with some
red/green color liquid. Wet gas contains less m e t h a n e t h a n a
dry gas does, b u t a larger fraction of C2-C 6 c o m p o n e n t s . Ob-
viously the m a i n difference between these reservoir fluids is
their respective composition. An example of c o m p o s i t i o n of
different reservoir fluids is given i n Table 1.2.
I n Table 1.2, C7+ refers to all h y d r o c a r b o n s having seven
or higher c a r b o n atoms a n d is called heptane-plus fraction,
while C6 refers to a group of all h y d r o c a r b o n s with six car-
b o n atoms (hexanes) that exist in the fluid. MT+ a n d SG7+ are
the molecular weight a n d specific gravity at 15.5~ (60~ for
the C7+ fraction of the mixture, respectively. It should be re-
alized that molecular weight a n d specific gravity of the whole
reservoir fluid are less t h a n the corresponding values for the

TABLE 1.2---Composition (mol%) and properties of various reservoir fluids and a crude oil
Component Dry gas~ Wet gasb Gas condensateC Volatileoild Blackoile Crudeoilf
CO2 3.70 0.00 0.18 1.19 0.09 0.00
N2 0.30 0.00 0.13 0.51 2.09 0.00
H2S 0.00 0.00 0.00 0.00 1.89 0.00
C1 96.00 82.28 61.92 45.21 29.18 0.00
C2 0.00 9.52 14.08 7.09 13.60 0.19
C3 0.00 4.64 8.35 4.61 9.20 1.88
iC4 0.00 0.64 0.97 1.69 0.95 0.62
nC4 0.00 0.96 3.41 2.81 4.30 3.92
iC5 0.00 0.35 0.84 1.55 1.38 2.11
nC5 0.00 0.29 1.48 2.01 2.60 4.46
C6 0.00 0.29 1.79 4.42 4.32 8.59
C7+ 0.00 1.01 6.85 28.91 30.40 78.23
Total 100.00 100.00 100.00 100.00 100.00 100.00
GOR (scf/stb) ... 69917 4428 1011 855
M7+ ... 113 143 190 209.8 266
SG7+ (at 15.5~ ... 0.794 0.795 0.8142 0.844 0.895
API7+ 46.7 46.5 42.1 36.1 26.6
"Gas sample from Salt Lake, Utah [12].
bWetgas data from McCaln[11].
CGas condensate sample from Samson County,Texas (M. B. Standing, personal notes, Department of Petroleum
Engineering,NorwegianInstitute of Technology,Trondheim,Norway, 1974).
dVolatile oil sample from Raleigh Field, Smith County,Mississipi(M. B. Standing,personal notes, Departmentof
Petroleum Engineering,NorwegianInstitute of Technology,Trondheim,Norway, 1974).
eBlack oil sample from M. Ghuraiba,M.Sc. Thesis, Kuwait University,Kuwait, 2000.
fA crude oil sampleproduced at stock tank conditions.

heptane-plus fraction. For example, for the crude oil sample
in Table 1.2, the specific gravity of the whole crude oil is 0.871
or API gravity of 31. Details of such calculations are discussed
in Chapter 4. These compositions have been determined from
recombination of the compositions of corresponding sepa-
rator gas and stock tank liquid, which have been measured
through analytical tools (i.e., gas chromatography, mass spec-
trometry, etc.). Composition of reservoir fluids varies with the
reservoir pressure and reservoir depth. Generally in a produc-
ing oil field, the sulfur and amount of heavy compounds in-
crease versus production time [10]. However, it is important
to note that within an oil field, the concentration of light hy-
drocarbons and the API gravity of the reservoir fluid increase
with the reservoir depth, while its sulfur and C7+ contents de-
crease with the depth [ 1]. The lumped C7+ fraction in fact is
a mixture of a very large number of hydrocarbons, up to C40
or higher. As an example the number of pure hydrocarbons
from C5 to C9 detected by chromatography tools in a crude oil
from North Sea reservoir fluids was 70 compounds. Detailed
composition of various reservoir fluids from the North Sea
fields is provided by Pedersen et al. [13]. As shown in Chapter
9, using the knowledge of the composition of a reservoir fluid,
one can determine a pressure-temperature (PT) diagram of
the fluid. And on the basis of the temperature and pressure
of the reservoir, the exact type of the reservoir fluid can be
determined from the PT diagram.
Reservoir fluids from a producing well are conducted to
two- or three-stage separators which reduce the pressure and
temperature of the stream to atmospheric pressure and tem-
perature. The liquid leaving the last stage is called stock tank
oil (STO) and the gas released in various stages is called as-
sociated gas. The liquid oil after necessary field processing is
called crude oil. The main factor in operation and design of an
oil-gas separator is to find the optimum operating conditions
of temperature and pressure so that the amount of produced
liquid (oil) is maximized. Such conditions can be determined
through phase behavior calculations, which are discussed in
detail in Chapter 9. Reservoir fluids from producing wells are
mixed with free water. The water is separated through gravi-
tational separators based on the difference between densities
of water and oil. Remaining water from the crude can be re-
moved through dehydration processes. Another surface oper-
ation is the desalting process that is necessary to remove the
salt content of crude oils. Separation of oil, gas, and water
from each other and removal of water and salt from oil and
any other process that occurs at the surface are called surface
production operations [14].
The crude oil produced from the atmospheric separator has
a composition different from the reservoir fluid from a pro-
ducing well. The light gases are separated and usually crude
oils have almost no methane and a small C2-C3 content while
the C7+ content is higher than the original reservoir fluid. As
an example, the composition of a crude oil produced through
a three-stage separator from a reservoir fluid is also given in
Table 1.2. Actually this crude is produced from a black oil
reservoir fluid (composition given in Table 1.2). Two impor-
tant characterisitcs of a crude that determine its quality are
the API gravity (specific gravity) and the sulfur content. Gen-
erally, a crude with the API gravity of less than 20 (SG > 0.934)
is called heavy crude and with API gravity of greater than 40
(SG < 0.825) is called light crude [1, 9]. Similarly, if the sulfur
1. I N T R O D U C T I O N 7
content of a crude is less than 0.5 wt% it is called a sweet
oil. It should be realized that these ranges for the gravity and
sulfur content are relative and may vary from one source to
another. For example, Favennec [15] classifies heavy crude as
those with API less than 22 and light crude having API above
33. Further classification of crude oils will be discussed in
Chapter 4.
1.1.3 P e t r o l e u m Fractions and Products
A crude oil produced after necessary field processing and
surface operations is transferred to a refinery where it is
processed and converted into various useful products. The
refining process has evolved from simple batch distillation
in the late nineteenth century to today's complex processes
through modern refineries. Refining processes can be gener-
ally divided into three major types: (1) separation, (2) con-
version, and (3) finishing. Separation is a physical process
where compounds are separated by different techniques. The
most important separation process is distillation that occurs
in a distillation column; compounds are separated based on
the difference in their boiling points. Other major physical
separation processes are absorption, stripping, and extrac-
tion. In a gas plant of a refinery that produces light gases,
the heavy hydrocarbons (Cs and heavier) in the gas mixture
are separated through their absorption by a liquid oil sol-
vent. The solvent is then regenerated in a stripping unit. The
conversion process consists of chemical changes that occur
with hydrocarbons in reactors. The purpose of such reactions
is to convert hydrocarbon compounds from one type to an-
other. The most important reaction in m o d e m refineries is
the cracking in which heavy hydrocarbons are converted to
lighter and more valuable hydrocarbons. Catalytic cracking
and thermal cracking are commonly used for this purpose.
Other types of reactions such as isomerization or alkylation
are used to produce high octane number gasoline. Finishing is
the purification of various product streams by processes such
as desulfurization or acid treatment of petroleum fractions to
remove impurities from the product or to stabilize it.
After the desalting process in a refinery, the crude oil en-
ters the atmospheric distillation column, where compounds
are separated according to their boiling points. Hydrocarbons
in a crude have boiling points ranging from -160~ (boil-
ing point of methane) to more than 600~ (ll00~ which
is the boiling point of heavy compounds in the crude oil.
However, the carbon-carbon bond in hydrocarbons breaks
down at temperatures around 350~ (660~ This process is
called cracking and it is undesirable during the distillation
process since it changes the structure of hydrocarbons. For
this reason, compounds having boiling points above 350~
(660+~ called residuum are removed from the bottom of
atmospheric distillation column and sent to a vacuum dis-
tillation column. The pressure in a vacuum distillation col-
umn is about 50-100 m m Hg, where hydrocarbons are boiled
at much lower temperatures. Since distillation cannot com-
pletely separate the compounds, there is no pure hydrocarbon
as a product of a distillation column. A group of hydrocarbons
can be separated through distillation according to the boiling
point of the lightest and heaviest compounds in the mixtures,
The lightest product of an atmospheric column is a mixture of
methane and ethane (but mainly ethane) that has the boiling
8 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

TABLE 1.3---Somepetroleum fractions produced from distillation columns.

Approximateboilingrange
Petroleum fraction Approximatehydrocarbonrange ~ ~
Light gases C2-C4 -90 to 1 -130-30
Gasoline (light and heavy) C4-C10 -1-200 30-390
Naphthas (light and heavy) C4-Cll -1-205 30-400
Jet fuel C9-C14 150-255 300-490
Kerosene C11-C14 205-255 400-490
Diesel fuel C]1-C16 205-290 400-550
Light gas oil C14-C18 255-315 490-600
Heavy gas oil C18-C28 315-425 600-800
Wax Cls-Ca6 315-500 600-930
Lubricating oil >C25 >400 >750
Vacuum gas oil C28-C55 425-600 800-1100
Residuum > C55 > 600 > 1100
Information given in this table is obtained from different sources [ 1,18,19].

range of -180 to -80~ ( - 2 6 0 to -40~ which corresponds molecular weight multiring aromatics. The vacuum residuum
to the boiling point of methane and ethane. This mixture, may be mixed with lighter products to produce a more valu-
which is in the form of gas and is known as fuel gas, is actu- able blend.
ally a petroleum fraction. In fact, during distillation a crude Distillation of a crude oil can also be performed in the lab-
is converted into a series of petroleum fractions where each oratory to divide the mixture into many narrow boiling point
one is a mixture of a limited number of hydrocarbons with a range fractions with a boiling range of about 10~ Such nar-
specific range of boiling point. Fractions with a wider range row range fractions are sometimes referred to as petroleum
of boiling points contain greater numbers of hydrocarbons. cuts. When boiling points of all the cuts in a crude are known,
All fractions from a distillation column have a known boiling then the boiling point distribution (distillation curve) of the
range, except the residuum for which the upper boiling point
is usually not known. The boiling point of the heaviest com-
ponent in a crude oil is not really known, but it is quite high. 60
The problem of the nature and properties of the heaviest com-
pounds in crude oils and petroleum residuum is still under
investigation by researchers [i 6, 17]. Theoretically, it can be
assumed that the boiling point of the heaviest component in a
Residuum
crude oil is infinity. Atmospheric residue has compounds with 50
.655
carbon number greater than 25, while vacuum residue has
compounds with carbon number greater than 50 (M > 800).
Some of the petroleum fractions produced from distillation
Vacuum
columns with their boiling point ranges and applications are 40 Gas Oil
given in Table 1.3. The boiling point and equivalent carbon
number ranges given in this table are approximate and they
may vary according to the desired specific product. For ex- 9
ample, the light gases fraction is mainly a mixture of ethane, E t-
propane, and butane; however, some heavier compounds zt- 30 fl_
(C5+) may exist in this fraction. The fraction is further frac- o Heavy t-
Gas Oil
tionated to obtain ethane (a fuel gas) and propane and butane A t m o s p h e r i c Distillation 46.1% o
(petroleum gases). The petroleum gases are liquefied to get
liquefied petroleum gas (LPG) used for home cooking pur- 345
Light
poses. In addition the isobutane may be separated for the gas 20
mixture to be used for improving vapor pressure characteris- ,205
Kerosene Gas Oil
tics (volatility) of gasoline in cold weathers. These fractions
may go through further processes to produce desired prod-
ucts. For example, gas oil may go through a cracking process 10-
to obtain more gasoline. Since distillation is not a perfect sep-
aration process, the initial and final boiling points for each V a c u u m Distillation 53.9o'/o
fraction are not exact and especially the end points are ap- ,q
proximate values. Fractions may be classified as narrow or t 90
wide depending on their boiling point range. As an example, 0 i i

the composition of an Alaska crude oil for various products 0 20 40 60 80 100

is given in Table 1.4 and is graphically shown in Fig. 1.3. Volume Percent
The weight and volume percentages for the products are
near each other. More than 50% of the crude is processed LightGases ~ LightGaserine
in vacuum distillation unit. The vacuum residuum is mainly
resin and asphaltenes-type compounds composed of high FIG. 1.3--Products and composition of Alaska crude oil.
 1. INTRODUCTION 9

TABLE 1.4--Products and composition of alaska crude oil.

Approximateboilingrangea
Petroleum fraction Approximatehydrocarbonrange ~ ~ vol% wt%
Atmospheric distillation
Light gases C2-C4 --90 to 1 --130-30 1.2 0.7
Light gasoline C4-C7 -1-83 30-180 4.3 3.5
Naphthas C7-Cll 83--205 180--400 16.0 14.1
Kerosene Cll-C16 205-275 400-525 12.1 11.4
Light gas oil (LGO) C16-C21 275-345 525-650 12.5 12.2
Sum C2-C21 -90-345 -130-650 46.1 41.9
Vacuum distillation (VD)
Heavy gas oil (HGO) C21-C31 345-455 650-850 20.4 21.0
Vacuum gas oil (VGO) C3l-C48 455-655 850-1050 15.5 16.8
Residuum >C48 655+ 1050+ 18.0 20.3
Sum C21-C48+ 345-655+ 650-1050 53.9 58.1
Total Crude C2-C48+ -9(P655+ -130-650+ 100.0 100.0
Informationgivenin this table has been extracted fromRef. [19].
aBoilingranges are interconvertedto the nearest 5~ (~

whole crude can be obtained. Such distillation data and their
uses will be discussed in Chapters 3 and 4. In a petroleum
cut, hydrocarbons of various types are lumped together in
four groups of paraffins (P), olefins (O), naphthenes (N), and
aromatics (A). For olefin-free petroleum cuts the composi-
tion is represented by the PNA content. If the composition
of a hydrocarbon mixture is known the mixture is called a
defined mixture, while a petroleum fraction that has an un-
known composition is called an undefined fraction.
As mentioned earlier, the petroleum fractions presented
in Table 1.3 are not the final products of a refinery. They
go through further physicochemical and finishing processes
to get the characteristics set by the market and government
regulations. After these processes, the petroleum fractions
presented in Table 1.3 are converted to petroleum products.
The terms petroleum fraction, petroleum cut, and petroleum
product are usually used incorrectly, while one should re-
alize that petroleum fractions are products of distillation
columns in a refinery before being converted to final prod-
ucts. Petroleum cuts may have very narrow boiling range
which may be produced in a laboratory during distillation
of a crude. In general the petroleum products can be divided
into two groups: (1) fuel products and (2) nonfuel products.
The major fuel petroleum products are as follows:
1. Liquefied petroleum gases (LPG) that are mainly used for
domestic heating and cooking (50%), industrial fuel (clean
fuel requirement) (15%), steam cracking feed stock (25%),
and as a motor fuel for spark ignition engines (10%). The
world production in 1995 was 160 million ton per year
(---5 million bbl/d) [20]. LPG is basically a mixture of
propane and butane.
2. Gasoline is perhaps one of the most important products of
a refinery. It contains hydrocarbons from C4 to Cll (molec-
ular weight of about 100-110). It is used as a fuel for cars.
Its main characteristics are antiknock (octane number),
volatility (distillation data and vapor pressure), stability,
and density. The main evolution in gasoline production has
been the use of unleaded gasoline in the world and the use
of reformulated gasoline (RFG) in the United States. The
RFG has less butane, less aromatics, and more oxygenates.
The sulfur content of gasoline should not exceed 0.03% by
weight. Further properties and characteristics of gasoline
will be discussed in Chapter 3. The U.S. gasoline demand
in 1964 was 4,4 million bbl/d and has increased from 7.2 to
8.0 million bbl/d in a period of 7 years from 1991 to 1998
[6, 20]. In 1990, gasoline was about a third of refinery prod-
ucts in the United States.
3. Kerosene and jet fuel are mainly used for lighting and jet
engines, respectively. The main characteristics are sulfur
content, cold resistance (for jet fuel), density, and ignition
quality,
4. Diesel and heating oil are used for motor fuel and domestic
purposes. The main characteristics are ignition (for diesel
oil), volatility, viscosity, cold resistance, density, sulfur con-
tent (corrosion effects), and flash point (safety factor).
5. Residual fuel oil is used for industrial fuel, for thermal pro-
duction of electricity, and as motor fuel (low speed diesel
engines). Its main characteristics are viscosity (good at-
omization for burners), sulfur content (corrosion), stabil-
ity (no decantation separation), cold resistance, and flash
point for safety.
The major nonfuel petroleum products are [18] as follows:
i. Solvents are light petroleum cuts in the C4-C14 range and
have numerous applications in industry and agriculture.
As an example of solvents, white spirits which have boiling
points between 135 and 205~ are used as paint thinners.
The main characteristics of solvents are volatility, purity,
odor, and toxicity. Benzene, toluene, and xylenes are used
as solvents for glues and adhesives and as a chemical for
petrochemical industries.
2. Naphthas constitute a special category of petroleum sol-
vents whose boiling points correspond to the class of white
spirits. They can be classified beside solvents since they are
mainly used as raw materials for petrochemicals and as
the feeds to steam crackers. Naphthas are thus industrial
intermediates and not consumer products. Consequently,
naphthas are not subject to government specifications but
only to commercial specifications.
3. Lubricants are composed of a main base stock and addi-
tives to give proper characteristics. One of the most im-
portant characteristics of lubricants is their viscosity and
viscosity index (change of viscosity with temperature). Usu-
ally aromatics are eliminated from lubricants to improve
10 CHARACTERIZATION AND P R O P E R T I E S OF P E T R O L E U M FRACTIONS
their viscosity index. Lubricants have structure similar
to isoparaffinic compounds. Additives used for lubricants
are viscosity index additives such as polyacrylates and
olefin polymers, antiwear additives (i.e., fatty esters), an-
tioxidants (i.e., alkylated aromatic amines), corrosion in-
hibitors (i.e., fatty acids), and antifoaming agents (i.e., poly-
dimethylsiloxanes). Lubricating greases are another class
of lubricants that are semisolid. The properties of lubri-
cants that should be known are viscosity index, aniline
point (indication of aromatic content), volatility, and car-
bon residue.
4. Petroleum waxes are of two types: the paraffin waxes in
petroleum distillates and the microcrystalline waxes in pe-
troleum residua. In some countries such as France, paraf-
fin waxes are simply called paraffins. Paraffin waxes are
high melting point materials used to improve the oil's pour
point and are produced during dewaxing of vacuum dis-
tillates. Paraffin waxes are mainly straight chain alkanes
(C18-C36) with a very small proportion of isoalkanes and
cycloalkanes. Their freezing point is between 30 and 70~
and the average molecular weight is around 350. When
present, aromatics appear only in trace quantities. Waxes
from petroleum residua (microcrystalline form) are less
defined aliphatic mixtures of n-alkanes, isoalkanes, and cy-
cloalkanes in various proportions. Their average molecular
weights are between 600 and 800, carbon number range is
a l k a n e s C30-C60 , and the freezing point range is 60-90~
[ 13]. Paraffin waxes (when completely dearomatized) have
applications in the food industry and food packaging. They
are also used in the production of candles, polishes, cos-
metics, and coatings [ 18]. Waxes at ordinary temperature of
25~ are in solid states although they contain some hydro-
carbons in liquid form. When melted they have relatively
low viscosity.
5. Asphalt is another major petroleum product that is pro-
duced from vacuum distillation residues. Asphalts contain
nonvolatile high molecular weight polar aromatic com-
pounds, such as asphaltenes (molecular weights of several
thousands) and cannot be distilled even under very high
vacuum conditions. In some countries asphalt is called
bitumen, although some suggest these two are different
petroleum products. Liquid asphaltic materials are in-
tended for easy applications to roads. Asphalt and bitu-
men are from a category of products called hydrocarbon
binders. Major properties to determine the quality of as-
phalt are flash point (for safety), composition (wax con-
tent), viscosity and softening point, weathering, density or
specific gravity, and stability or chemical resistance.
6. There are some other products such as white oils (used in
pharmaceuticals or in the food industry), aromatic extracts
(used in the paint industry or the manufacture of plastics),
and coke (as a fuel or to produce carbon elecrodes for alu-
m i n u m refining). Petroleum cokes generally have boiling
points above 1100+~ (~2000+~ molecular weight of
above 2500+, and carbon number of above 200+. Aromatic
extracts are black materials, composed essentially of con-
densed polynuclear aromatics and of heterocyclic nitrogen
and/or sulfur compounds. Because of this highly aromatic
structure, the extracts have good solvent power.
Further information on technology, properties, and test-
ing methods of fuels and lubricants is given in Ref. [21].
In general, more than 2000 petroleum products within some
20 categories are produced in refineries in the United States
[ 1, 19]. Blending techniques are used to produce some of these
products or to improve their quality. The product specifica-
tions must satisfy customers' requirements for good perfor-
mance and government regulations for safety and environ-
ment protection. To be able to plan refinery operations, the
availability of a set of product quality prediction methods is
therefore very important.
There are a number of international organizations that are
known as standard organizations that recommend specific
characteristics or standard measuring techniques for various
petroleum products through their regular publications. Some
of these organizations in different countries that are known
with their abbreviations are as follows:
1. ASTM (American Society for Testing and Materials) in the
United States
2. ISO (International Organization for Standardization),
which is at the international level
3. IP (Institute of Petroleum) in the United Kingdom
4. API (American Petroleum Institute) in the United States
5. AFNOR (Association Francaise de Normalisation), an offi-
cial standard organization in France
6. Deutsche Institut fur Norrnung (DIN) in Germany
7. Japan Institute of Standards (J-IS) in Japan
ASTM is composed of several committees in which the D-02
committee is responsible for petroleum products and lubri-
cants, and for this reason its test methods for petroleum ma-
terials are designated by the prefix D. For example, the test
method ASTM D 2267 provides a standard procedure to de-
termine the benzene content of gasoline [22]. In France this
test method is designated by EN 238, which are documented
in AFNOR information document M 15-023. Most standard
test methods in different countries are very similar in prac-
tice and follow ASTM methods but they are designated by
different codes. For example the international standard ISO
6743/0, accepted as the French standard NF T 60-162, treats
all the petroleum lubricants, industrial oils, and related prod-
ucts. The abbreviation NF is used for the French standard,
while EN is used for European standard methods [ 18].
Government regulations to protect the environment or to
save energy, in m a n y cases, rely on the recommendations
of official standard organizations. For example, in France,
AFNOR gives specifications and requirements for various
petroleum products. For diesel fuels it recommends (after
1996) that the sulfur content should not exceed 0.05 wt% and
the flash point should not be less than 55~ [18].
1.2 T Y P E S A N D IMPORTANCE
OF PHYSICAL P R O P E R T I E S
On the basis of the production and refining processes de-
scribed above it may be said that the petroleum industry
is involved with m a n y types of equipment for production,
transportation, and storage of intermediate or final petroleum
products. Some of the most important units are listed below.
i. Gravity decanter (to separate oil and water)
2. Separators to separate oil and gas
3. Pumps, compressors, pipes, and valves

4. Storage tanks
5. Distillation, absorption, and stripping columns
6. Boilers, evaporators, condensers, and heat exchangers
7. Flashers (to separate light gases from a liquid)
8. Mixers and agitators
9. Reactors (fixed and fluidized beds)
10. Online analyzers (to monitor the composition)
11. Flow and liquid level measurement devices
12. Control units and control valves
The above list shows some, but not all, of the units involved
in the petroleum industry. Optimum design and operation
of such units as well as manufacture of products to meet
market demands and government regulations require a com-
plete knowledge of properties and characteristics for hydro-
carbons, petroleum fractions/products, crude oils, and reser-
voir fluids. Some of the most important characteristics and
properties of these fluids are listed below with some exam-
ples for their applications. They are divided into two groups
of temperature-independent parameters and temperature-
dependent properties. The temperature-independent properties
and parameters are as follows:
1. Specific gravity (SG) or density (d) at SC. These para-
meters are temperature-dependent; however, specific
gravity at 15.5~ and 1 atm and density at 20~ and 1
atm used in petroleum characterization are included in
this category of temperature-independent properties. The
specific gravity is also presented in terms of API gravity.
It is a useful parameter to characterize petroleum fluids,
to determine composition (PNA) and the quality of a fuel
(i.e., sulfur content), and to estimate other properties such
as critical constants, density at various temperatures, vis-
cosity, or thermal conductivity [23, 24]. In addition to its
direct use for size calculations (i.e., pumps, valves, tanks,
and pipes), it is also needed in design and operation of
equipments such as gravity decanters.
2. Boiling point (Tb) or distillation curves such as the true
boiling point curve of petroleum fractions. It is used to
determine volatility and to estimate characterization pa-
rameters such as average boiling point, molecular weight,
composition, and many physical properties (i.e., critical
constants, vapor pressure, thermal properties, transport
properties) [23-25].
3. Molecular weight (M) is used to convert molar quantities
into mass basis needed for practical applications. Ther-
modynamic relations always produce molar quantities
(i.e., molar density), while in practice mass specific val-
ues (i.e., absolute density) are needed. Molecular weight
is also used to characterize oils, to predict composition
and quality of oils, and to predict physical properties such
as viscosity [26-30].
4. Refractive index (n) at some reference conditions (i.e., 20~
and 1 atm) is another useful characterization parameter
to estimate the composition and quality of petroleum frac-
tions. It is also used to estimate other physical properties
such as molecular weight, equation of state parameters,
the critical constants, or transport properties of hydrocar-
bon systems [30, 31].
5. Defined characterization parameters such as Watson K,
carbon-to-hydrogen weight ratio, (CH weight ratio), refrac-
tivity intercept (Ri), and viscosity gravity constant (VGC)
1. I N T R O D U C T I O N 11
to determine the quality and composition of petroleum
fractions [27-29].
6. Composition of petroleum fractions in terms of wt% of
paraffins (P%), naphthenes (N%), aromatics (A%), and
sulfur content (S%) are important to determine the qual-
ity of a petroleum fraction as well as to estimate physical
properties through pseudocomponent methods [31-34].
Composition of other constituents such as asphaltene and
resin components are quite important for heavy oils to
determine possibility of solid-phase deposition, a major
problem in the production, refining, and transportation
of oil [35].
7. Pour point (Tp), and melting point (TM) have limited uses
in wax and paraffinic heavy oils to determine the degree
of solidification and the wax content as well as m i n i m u m
temperature required to ensure fluidity of the oil.
8. Aniline point to determine a rough estimate of aromatic
content of oils.
9. Flash point (TF) is a very useful property for the safety of
handling volatile fuels and petroleum products especially
in summer seasons.
10. Critical temperature (To), critical pressure (Pc), and critical
volume (Vc) known as critical constants or critical prop-
erties are used to estimate various physical and thermo-
dynamic properties through equations of state or gener-
alized correlations [36].
11. Acentric factor (w) is another parameter that is needed
together with critical properties to estimate physical and
thermodynamic properties through equations of state
[36].
The above properties are mainly used to characterize the
oil or to estimate the physical and thermodynamic proper-
ties which are all temperature-dependent. Some of the most
important properties are listed as follows:
1. Density (p) as a function of temperature and pressure
is perhaps the most important physical property for
petroleum fluids (vapor or liquid forms). It has great ap-
plication in both petroleum production and processing as
well as its transportation and storage. It is used in the
calculations related to sizing of pipes, valves, and storage
tanks, power required by pumps and compressors, and
flow-measuring devices. It is also used in reservoir simula-
tion to estimate the amount of oil and gas in a reservoir, as
well as the amount of their production at various reservoir
conditions. In addition density is used in the calculation
of equilibrium ratios (for phase behavior calculations) as
well as other properties, such as transport properties.
2. Vapor pressure (pv~p) is a measure of volatility and it is
used in phase equilibrium calculations, such as flash, bub-
ble point, or dew point pressure calculations, in order to
determine the state of the fluid in a reservoir or to sep-
arate vapor from liquid. It is needed in calculation of
equilibrium ratios for operation and design of distilla-
tion, absorber, and stripping columns in refineries. It is
also needed in determination of the amount of hydrocar-
bon losses from storage facilities and their presence in
air. Vapor pressure is the property that represents igni-
tion characteristics of fuels. For example, the Reid vapor
pressure (RVP) and boiling range of gasoline govern ease
of starting engine, engine warm-up, rate of acceleration,
mileage economy, and tendency toward vapor lock [ 19].
12 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
3. Heat capacity (Cp) of a fluid is needed in design and oper-
ation of heat transfer units such as heat exchangers.
4. Enthalpy (H) of a fluid is needed in energy balance cal-
culations, heat requirements needed in design and oper-
ation of distillation, absorption, stripping columns, and
reactors.
5. Heat of vaporization (AHvap) is needed in calculation of
heat requirements in design and operation of reboilers or
condensers.
6. Heats of formation (hHf), combustion (AHc), and reaction
(AHr) are used in calculation of heating values of fuels
and the heat required/generated in reactors and furnaces
in refineries. Such information is essential in design and
operations of burners, furnaces, and chemical reactors.
These properties together with the Gibbs free energy are
used in calculation of equilibrium constants in chemical
reactions to determine the optimum operating conditions
in reactors for best conversion of feed stocks into the prod-
ucts.
7. Viscosity (t*) is another useful property in petroleum pro-
duction, refining, and transportation. It is used in reser-
voir simulators to estimate the rate of oil or gas flow
and their production. It is needed in calculation of power
required in mixers or to transfer a fluid, the amount of
pressure drop in a pipe or column, flow measurement de-
vices, and design and operation of oil/water separators
[37, 38].
8. Thermal conductivity (k) is needed for design and opera-
tion of heat transfer units such as condensers, heat ex-
changers, as well as chemical reactors [39].
9. Diffusivity or diffusion coefficient (D) is used in calcula-
tion of mass transfer rates and it is a useful property in
design and operation of reactors in refineries where feed
and products diffuse in catalyst pores. In petroleum pro-
duction, a gas injection technique is used in improved oil
recovery where a gas diffuses into oil under reservoir con-
ditions; therefore, diffusion coefficient is also required in
reservoir simulation and modeling [37, 40-42].
10. Surface tension (a) or interfacial tension (IFT) is used
mainly by the reservoir engineers in calculation of cap-
illary pressure and rate of oil production and is needed
in reservoir simulators [37]. In refineries, IFT is a use-
ful parameter to determine foaming characteristics of oils
and the possibility of having such problems in distillation,
absorption, or stripping columns [43]. It is also needed
in calculation of the rate of oil dispersion on seawater
surface polluted by an oil spill [44].
11. Equilibrium ratios (Ki) and fugacity coefficients (~Pi) are
the most important thermodynamic properties in all
phase behavior calculations. These calculations include
vapor-liquid equilibria, bubble and dew point pressure,
pressure-temperature phase diagram, and GOR. Such cal-
culations are important in design and operation of distilla-
tion, absorption and stripping units, gas-processing units,
gas-oil separators at production fields, and to determine
the type of a reservoir fluid [45, 46].
Generally, the first set of properties introduced above
(temperature-independent) are the basic parameters that are
used to estimate physical and thermodynamic properties
given in the second set (temperature-dependent). Properties
such as density, boiling point, molecular weight, and refrac-
tive index are called physical properties. Properties such as en-
thalpy, heat capacity, heat of vaporization, equilibrium ratios,
and fugacity are called thermodynamic properties. Viscosity,
thermal conductivity, diffusion coefficient, and surface ten-
sion are in the category of physical properties but they are also
called transport properties. In general all the thermodynamic
and physical properties are called thermophysical properties.
But they are commonly referred to as physical properties or
simply properties, which is used in the title of this book.
A property of a system depends on the thermodynamic state
of the system that is determined by its temperature, pressure,
and composition. A process to experimentally determine var-
ious properties for all the industrially important materials,
especially complex mixtures such as crude oils or petroleum
products, would be prohibitive in both cost and time, indeed
it could probably never be completed. For these reasons ac-
curate methods for the estimation of these properties are be-
coming increasingly important. In some references the term
property prediction is used instead of property estimation;
however, in this book as generally adopted by most scientists
both terms are used for the same purpose.
1.3 I M P O R T A N C E OF P E T R O L E U M F L U I D S
CHARACTERIZATION
In the previous section, various basic characteristic para-
meters for petroleum fractions and crude oils were intro-
duced. These properties are important in design and oper-
ation of almost every piece of equipment in the petroleum in-
dustry. Thermodynamic and physical properties of fluids are
generally calculated through standard methods such as cor-
responding state correlations or equations of state and other
pressure-volume-temperature (PVT) relations. These corre-
lations and methods have a generally acceptable degree of ac-
curacy provided accurate input parameters are used. When
using cubic equation of state to estimate a thermodynamic
property such as absolute density for a fluid at a known tem-
perature and pressure, the critical temperature (Tc), critical
pressure (Pc), acentric factor (~0), and molecular weight (M)
of the system are required. For most pure compounds and hy-
drocarbons these properties are known and reported in var-
ious handbooks [36, 47-50]. If the system is a mixture such
as a crude oil or a petroleum fraction then the pseudocritical
properties are needed for the calculation of physical proper-
ties. The pseudocritical properties cannot be measured but
have to be calculated through the composition of the mix-
ture. Laboratory reports usually contain certain measured
properties such as distillation curve (i.e., ASTM D 2887) and
the API gravity or specific gravity of the fraction. However,
in some cases viscosity at a certain temperature, the per-
cent of paraffin, olefin, naphthene, and aromatic hydrocar-
bon groups, and sulfur content of the fraction are measured
and reported. Petroleum fractions are mixtures of many com-
pounds in which the specific gravity can be directly measured
for the mixture, but the average boiling point cannot be mea-
sured. Calculation of average boiling point from distillation
data, conversion of various distillation curves from one type
to another, estimation of molecular weight, and the PNA com-
position of fractions are the initial steps in characterization of

petroleum fractions [25, 46, 47]. Estimation of other basic pa-
rameters introduced in Section 1.2, such as asphaltenes and
sulfur contents, CH, flash and pour points, aniline point, re-
fractive index and density at SC, pseudocrtitical properties,
and acentric factor, are also considered as parts of charac-
terization of petroleum fractions [24, 28, 29, 51-53]. Some of
these properties such as the critical constants and acentric
factor are not even known for some heavy pure hydrocarbons
and should be estimated from available properties. Therefore
characterization methods also apply to pure hydrocarbons
[33]. Through characterization, one can estimate the basic
parameters needed for the estimation of various physical and
thermodynamic properties as well as to determine the com-
position and quality of petroleum fractions from available
properties easily measurable in a laboratory.
For crude oils and reservoir fluids, the basic laboratory
data are usually presented in the form of the composition
of hydrocarbons up to hexanes and the heptane-plus frac-
tion (C7+), with its molecular weight and specific gravity
as shown in Table 1.2. In some cases laboratory data on a
reservoir fluid is presented in terms of the composition of
single carbon numbers or simulated distillation data where
weight fraction of cuts with known boiling point ranges are
given. Certainly because of the wide range of compounds ex-
isting in a crude oil or a reservoir fluid (i.e., black oil), an
average value for a physical property such as boiling point
for the whole mixture has little significant application and
meaning. Characterization of a crude oil deals with use of
such laboratory data to present the mixture in terms of a
defined or a continuous mixture. One commonly used char-
acterization technique for the crudes or reservoir fluids is
to represent the hydrocarbon-plus fraction (C7+) in terms of
several narrow-boiling-range cuts called psuedocomponents
(or pseudofractions) with known composition and character-
ization parameters such as, boiling point, molecular weight,
and specific gravity [45, 54, 55]. Each pseudocomponent is
treated as a petroleum fraction. Therefore, characterization
of crude oils and reservoir fluids require characterization of
petroleum fractions, which in turn require pure hydrocarbon
characterization and properties [56]. It is for this reason that
properties of pure hydrocarbon compounds and hydrocarbon
characterization methods are first presented in Chapter 2,
the characterization of petroleum fractions is discussed in
Chapter 3, and finally methods of characterization of crude
oils are presented in Chapter 4. Once characterization of a
petroleum fraction or a crude oil is done, then a physical
property of the fluid can be estimated through an appropri-
ate procedure. In summary, characterization of a petroleum
fraction or a crude oil is a technique that through available
laboratory data one can calculate basic parameters necessary
to determine the quality and properties of the fluid.
Characterization of petroleum fractions, crude oils, and
reservoir fluids is a state-of-the-art calculation and plays an
important role in accurate estimation of physical properties
of these complex mixtures. Watson, Nelson, and Murphy of
Universal Oi1 Products (UOP) in the mid 1930s proposed ini-
tial characterization methods for petroleum fractions [57].
They introduced a characterization parameter known as
Watson or UOP characterization factor, Kw, which has been
used extensively in characterization methods developed in the
following years. There are many characterization methods
1. INTRODUCTION 13
30
r~
,<
<C
20
0
.~ e%Q
"0
10
_~.,~
\
-. :.,.. - - 9......... .i
: ,,,-.,......,.
" , , ~ x e ~ ..," o~"*~" ~ 6e~'*l'~9
I ,I I I i ,
-6 -4 -2 0 2 4
Deviation f r o m E x p e r i m e n t a l Value, %
FIG. 1.~ !nfluence of error in critical t e m p e r a t u r e on errors
in predicted physical properties of toluene. Taken from Ref. [58]
with p e r m i s s i o n .
suggested in the literature or in process simulators and each
method generates different characterization parameters that
in turn would result different estimated final physical prop-
erty with subsequent impact in design and operation of re-
lated units. To decide which method of characterization and
what input parameters (where there is a choice) should be
chosen depends very much on the user's knowledge and ex-
perience in this important area.
To show how important the role of characterization is in
the design and operation of units, errors in the prediction
of various physical properties of toluene through a modified
BWR equation of state versus errors introduced to actual crit-
ical temperature (To) are shown in Fig. 1.4 [58]. In this figure,
errors in the prediction of vapor pressure, liquid viscosity,
vapor viscosity, enthalpy, heat of vaporization, and liquid den-
sity are calculated versus different values of critical tempera-
ture while other input parameters (i.e., critical pressure, acen-
tric factor, etc.) were kept constant. In the use of the equation
of state if the actual (experimental) value of the critical tem-
perature is used, errors in values of predicted properties are
generally within 1-3% of experimental values; however, as
higher error is introduced to the critical temperature the error
in the calculated property increases to a much higher magni-
tude. For example, when the error in the value of the critical
temperature is zero (actual value of Tc), predicted vapor pres-
sure has about 3% error from the experimental value, but
when the error in Tc increases to 1, 3, or 5%, error in the pre-
dicted vapor pressure increases approximately to 8, 20, and
40%, respectively. Therefore, one can realize that 5% error in
an input property for an equation of state does not necessar-
ily reflect the same error in a calculated physical property but
can be propagated into much higher errors, while the pre-
dictive equation is relatively accurate if actual input parame-
ters are used. Similar results are observed for other physical
14 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
100
80 .....
x~(,~. -- --
60 "~ \ ' X .....
40 ~ n-Tetracosane
20 "%\
"~~'~~' -20.400 "~ " ~
-60
-6 -4 -2 0 2
% Deviationin CriticalTemperature
FIG. 1 . 5 ~ l n f l u e n c e of error in critical temperature on errors
of predicted vapor pressure from Lee-Kesler method.
properties and with other correlations for the estimation of
physical properties [59]. Effect of the error in the critical
temperature on the vapor pressure of different compounds
predicted from the Lee-Kesler method (see Section 7.3.2) is
shown in Fig. 1.5. When the actual critical temperature is
used, the error in the predicted vapor pressure is almost neg-
ligible; however, if the critical temperature is under-predicted
by 5%, the error in the vapor pressure increases by 60-80%
for the various compounds evaluated.
As shown in Chapter 6, vapor pressure is one of the key
parameters in the calculation of equilibrium ratios (Ki) and
subsequent relative volatility (a12), which is defined in a bi-
nary system of components 1 and 2 as follows:
(1.3) K1 = y~
X1
Ka Yl x - X2
(1.4) oq2 = ~ = x--~ Y2
where xl and x2 are the mole fractions of components 1 and
2 in the liquid phase, respectively. Similarly yt and y2 are the
mole fractions in the vapor phase for components 1 and 2,
respectively. For an ideal binary system at low pressure, the
equilibrium ratio Ki is directly proportional to the vapor pres-
sure as will be seen in Chapter 6.
The most important aspect in the design and operation of
distillation columns is the number or trays needed to make a
specific separation for specific feed and products. It has been
shown that a small error in the value of relative volatility could
lead to a m u c h greater error in the calculation of number of
trays and the length of a distillation column [60]. The mini-
m u m n u m b e r of trays required in a distillation column can be
calculated from the knowledge of relative volatility through
the Fenske Equation given below [61].
(1.5) Nmin = l n [ x D ( 1 -- X B ) / X B ( 1 - - XD)] -- 1
ln(ulz)
where Nmin is the m i n i m u m number of plates, and xD and xB
are the mole fraction of the light component in the distillate
n-Hexane (top) and bottom products, respectively. Equation (1.5) is de-
n-Octane veloped for a binary mixture; however, a similar equation has
n-Eieosane been developed for multicomponent mixtures [61]. For differ-
ent values of or, errors calculated for the m i n i m u m n u m b e r
of trays versus errors introduced in the value of ~ through
Eq. (1.5) are shown in Fig. 1.6. As is shown in this figure,
a - 5 % error in the value of a when its value is 1.1 can gen-
erate an error of more than 100% in the calculation of min-
i m u m number of trays. It can be imagined that the error in
the actual number of trays would be even higher than 100%.
In addition, the calculated numbers of trays are theoretical
and when converted to real number of trays through overall
column efficiency, the error may increase to several hundred
percent. The approach of building the column higher to have
4 6 a safe design is quite expensive.
As an example, a distillation column of diameter 4.5 m
and height 85 m has an investment cost of approximately
$4 million (~4.5 million) as stated by Dohrn and Pfohl [60].
Error in the calculation of relative volatility, a, could have
been caused by the error in calculation of vapor pressure,
which itself could have been caused by a small error in an
input parameter such as critical temperature [58, 59]. There-
fore, from this simple analysis one can realize the extreme
cost and loss in the investment that can be caused by a small
error in the estimation of critical temperature. Similar other
examples have been given in the literature [62]. Nowadays,
investment in refineries or their upgrading costs billions of
dollars. For example, for a typical refinery of 160000 bbl/d
(8 million tons/year) capacity, the cost of construction in
Europe is about $2 billion [18]. This is equivalent to refining
cost of $7.5/bbl while this number for refineries of 1980s
was about $2/bbl. In addition to the extra cost of investment,
inappropriate design of units can cause extra operating costs
and shorten the plant life as well as produce products that
do not match the original design specifications. The use of a
proper characterization method to calculate more accurate
100
80 a=l.l
60 ..... a=l.2
~) 4O a= 1.3
20
z
o
Ze -2o
-4o
4o
-80
-100
-10 -8 -6 -4 -2 0 2 4 6 g 10
% Error in RelativeVolatility
FIG. 1 . & - E f f e c t of error in the relative volatil-
ity on the error of minimum number of plates
of a distillation column.

properties of petroleum fractions can save a large portion of
such huge additional investment and operating costs.
1.4 O R G A N I Z A T I O N O F T H E B O O K
As the title of the book portrays and was discussed in Sec-
tions 1.2 and 1.3, the book presents methods of characteriza-
tion and estimation of thermophysical properties of hydrocar-
bons, defined mixtures, undefined petroleum fractions, crude
oils, and reservoir fluids. The entire book is written in nine
chapters in a way such that in general every chapter requires
materials presented in previous chapters. In addition there is
an appendix and an index. Chapter 1 gives a general intro-
duction to the subject from basic definition of various terms,
the nature of petroleum, its formation and composition, types
of petroleum mixtures, and the importance of characteriza-
tion and property prediction to specific features of the book
and its application in the petroleum industry and academia.
Because of the importance of units in property calculations,
the last section of Chapter 1 deals with unit conversion fac-
tors especially between SI and English units for the parame-
ters used in the book. Chapter 2 is devoted to properties and
characterization of pure hydrocarbons from C1 to C22 from
different hydrocarbon groups, especially from homologous
groups commonly found in petroleum fluids. Properties of
some nonhydrocarbons found with petroleum fluids such as
H20, H2S, CO2, and Nz are also given. Basic parameters are
defined at the beginning of the chapter, followed by charac-
terization of pure hydrocarbons. Predictive methods for vari-
ous properties of pure hydrocarbons are presented and com-
pared with each other. A discussion is given on the state-of-
the-arts development of predictive methods. The procedures
presented in this chapter are essential for characterization of
petroleum fractions and crude oils discussed in Chapters 3
and 4.
Chapter 3 discusses various characterization methods for
petroleum fractions and petroleum products. Characteriza-
tion parameters are introduced and analytical instruments in
laboratory are discussed. In this chapter one can use min-
i m u m laboratory data to characterize petroleum fractions
and to determine the quality of petroleum products. Esti-
mation of some basic properties such as molecular weight,
molecular-type composition, sulfur content, flash, pour point
and freezing points, critical constants, and acentric factor for
petroleum fractions are presented in this chapter. A theoret-
ical discussion on development of characterization methods
and generation of predictive correlations from experimental
data is also presented. Methods of Chapter 3 are extended to
Chapter 4 for the characterization of various reservoir fluids
and crude oils. Chapters 2-4 are perhaps the most impor-
tant chapters in this book, as the methods presented in these
chapters influence the entire field of physical properties in the
remaining chapters.
In Chapter 5, PVT relations, equations of state, and
corresponding state correlations are presented [31,63-65].
The use of the velocity of light and sound in developing
equations of state is also presented [31, 66-68]. Equations of
state and corresponding state correlations are powerful tools
in the estimation of volumetric, physical, transport, and
thermodynamic properties [64, 65, 69]. Procedures outlined
1. I N T R O D U C T I O N 15
in Chapter 5 will be used in the prediction of physical
properties discussed in the follow-up chapters. Fundamental
thermodynamic relations for calculation of thermodynamic
properties are presented in Chapter 6. The last three chapters
of the book show applications of methods presented in Chap-
ters 2-6 for calculation of various physical, thermodynamic,
and transport properties. Methods of calculation and esti-
mation of density and vapor pressure are given in Chapter 7.
Thermal properties such as heat capacity, enthalpy, heat
of vaporization, heats of combustion and reaction, and
the heating value of fuels are also discussed in Chapter 7.
Predictive methods for transport properties namely viscosity,
thermal conductivity, diffusixdty, and surface tension are
given in Chapter 8 [30,31,42,43,69,70]. Finally, phase
equilibrium calculations, estimation of equilibrium ratios,
GOR, calculation of pressure-temperature (PT) diagrams,
solid formations, the conditions at which asphaltene, wax,
and hydrate are formed, as well as their preventive methods
are discussed in Chapter 9.
The book is written according to the standards set by ASTM
for its publication. Every chapter begins with a general intro-
duction to the chapter. Since in the following chapters for
most properties several predictive methods are presented, a
section on conclusion and recommendations is added at the
end of the chapter. Practical problems as examples are pre-
sented and solved for each property discussed in each chap-
ter. Finally, the chapter ends by a set of exercise problems
followed by a citation section for the references used in the
chapter.
The Appendix gives a summary of definitions of terms and
properties used in this manual according to the ASTM dictio-
nary as well as the Greek letters used in this manual. Finally
the book ends with an index to provide a quick guide to find
specific subjects.
1.5 S P E C I F I C F E A T U R E S O F
THIS MANUAL
In this part several existing books in the area of character-
ization and physical properties of petroleum fractions are
introduced and their differences with the current book are
discussed. Then some special features of this book are pre-
sented.
1.5.1 Introduction of S o m e Existing Books
There are several books available that deal with physical prop-
erties of petroleum fractions and hydrocarbon systems. The
most comprehensive and widely used book is the API Tech-
nical Data Book--Petroleum Refining [47]. It is a book with
15 chapters in three volumes, and the first edition appeared
in mid 1960s. Every 5 years since, some chapters of the book
have been revised and updated. The project has been con-
ducted at the Pennsylvania State University and the sixth
edition was published in 1997. It contains a data bank on
properties of pure hydrocarbons, chapters on characteriza-
tion of petroleum fractions, thermodynamic and transport
properties of liquid and gaseous hydrocarbons, their mix-
tures, and undefined petroleum fractions. For each property,
one predictive method that has been approved and selected
16 CHARACTERIZATION AND P R O P E R T I E S OF P E T R O L E U M FRACTIONS
by the API-TDB committees as the best available method is
presented. This book will be referred as API-TDB throughout
this book.
Another important book in this area is The Properties of
Gases and Liquids that was originally written by Reid and
Sherwood in 1950s and it has been revised and updated nearly
every decade. The fifth and latest edition was published in
November 2000 [36] by three authors different from the orig-
inal two authors. The book has been an excellent reference for
students and practical engineers in the industry over the past
five decades. It discusses various methods for prediction of
properties of pure hydrocarbons as well as nonhydrocarbons
and their defined mixtures. However, it does not treat un-
defined petroleum fractions, crude oils, and reservoir fluids.
Most of the methods for properties of pure compounds re-
quire the chemical structure of compounds (i.e., group con-
tribution techniques). The book compares various methods
and gives its recommendations for each method.
There are several other books in the area of properties of
oils that document empirically developed predictive methods,
among them is the book Properties of Oils and Natural Gases,
by Pedersen et al. [ 13]. The book mainly treats reservoir fluids,
especially gas condensates from North Sea, and it is mainly
a useful reference for reservoir engineers. Books by McCain
[11], Ahmed [71], Whitson [45], and Danesh [72] are all writ-
ten by reservoir engineers and contain information mainly for
phase behavior calculations needed in petroleum production
and reservoir simulators. However, they contain some useful
information on methods of prediction of some physical prop-
erties of petroleum fractions. Another good reference book
was written by Tsonopoulos et al. [73] on thermodynamic
and transport properties of coal liquids in the mid 1980s.
Although there are m a n y similarities between coal liquids
and petroleum fractions, the book does not consider crude
oils and reservoir fluids. But it provides some useful correla-
tions for properties of coal liquids. The book by Wauquier [ 18]
on petroleum refining has several useful chapters on charac-
terization and physical properties of petroleum fractions and
finished products. It also provides the test methods accord-
ing to European standards. Some organizations' Web sites
also provide information on fluid physical properties. A good
example of such online information is provided by National
Institute of Standards (http://webbook.nist.gov) which gives
molecular weight, names, formulas, structure, and some data
on various compounds [74].
1.5.2 Special Features of the Book
This book has objectives and aims that are different from
the books mentioned in Section 1.5.1. The main objective
of this book has been to provide a quick reference in the
area of petroleum characterization and properties of various
petroleum fluids for the people who work in the petroleum
industry and research centers, especially in petroleum pro-
cessing (downstream), petroleum production (upstream), and
related industries. One special characteristic of the book is its
discussion on development of various methods which would
help the users of process/reservoir simulators to become fa-
miliar with the nature of characterization and property esti-
mation methods for petroleum fractions. This would in turn
help them to choose the proper predictive method among the
m a n y methods available in a process simulator. However, the
book has been written in a language that is understandable
to undergraduate and graduate students in all areas of engi-
neering and science. It contains practical solved problems as
well as exercise problems so that the book would be suitable
as a text for educational purposes.
Special features of this book are Chapters 2, 3, and 4 that
deal with the characterization of hydrocarbons, petroleum
fractions, and crude oils and their impact on the entire field
of property prediction methods. It discusses both light as
well as heavy fractions and presents methods of prediction
of the important characteristics of petroleum products from
m i n i m u m laboratory data and easily measurable parame-
ters. It presents several characterization methods developed
in recent years and not documented in existing references.
The book also presents various predictive methods, including
the most accurate and widely used method for each property
and discusses points of strength, weaknesses, and limitations.
Recommended methods are based on the generality, simplic-
ity, accuracy, and availability of input parameters. This is
another special feature of the book. In Chapters 5 and 6 it
discusses equations of state based on the velocity of sound
and light and how these two measurable properties can be
used to predict thermodynamic and volumetric properties of
fluids, especially heavy compounds and their mixtures [31,63,
66-68]. Significant attention is given throughout the book on
how to estimate properties of heavy hydrocarbons, petroleum
fractions, crude oils, and reservoir fluids. Most of the methods
developed by Riazi and coworkers [23, 24, 26-33, 51-56, 63,
65-70], which have been in use by the petroleum industry
[47, 75-82], are documented in this book. In addition, a new
experimental technique to measure diffusion coefficients in
reservoir fluids under reservoir conditions is presented in
Chapter 8 [42]. In Chapter 9 some new methods for determi-
nation of onset of solid formation are introduced. Reported
experimental data on characteristics and properties of var-
ious oils from different parts of the world are included in
various chapters for direct evaluations and testing of meth-
ods. Although both gases and liquids are treated in the book,
emphasis is on the liquid fractions. Generally, the methods
of estimation of properties of gases are more accurate than
those for liquid systems. Most of the methods presented in the
book are supported by some scientific basis and they are not
simply empirical correlations derived from a certain group of
data. This widens the application of the methods presented
in the book to different types of oils. However, all basic pa-
rameters and necessary engineering concepts are defined in
a way that is understandable for those nonengineer scientists
who are working in the petroleum or related industry. Nearly
all methods are expressed through mathematical relations so
they are convenient for computer applications; however, most
of them are simple such that the properties can be calculated
by hand calculators for a quick estimate whenever applica-
ble special methods are given for coal liquid fractions. This is
another unique feature of this book.
1.6 A P P L I C A T I O N S OF T H E B O O K
The information that is presented in the book may be applied
and used in all areas of the petroleum industries: production,

processing, and transportation. It can also be used as a
textbook for educational purposes. Some of the applica-
tions of the materials covered in the book were discussed in
Sections 1.2 and 1.3. The applications and uses of the book
may he summarized as follows.
1.6.1 A p p l i c a t i o n s in Petroleum Processing
(Downstream)
Engineers, scientists, and operators working in various sec-
tors of petroleum processing and refining or related industries
can use the entire material discussed in the book. It helps
laboratory people in refineries to measure useful properties
and to test the reliability of their measurements. The book
should be useful for engineers and researchers to analyze ex-
perimental data and develop their own predictive methods.
It is also intended to help people who are involved with de-
velopment of computer softwares and process simulators for
design and operation of units and equipments in petroleum
refineries. Another objective was to help users of such simu-
lators to be able to select an appropriate predictive method
for a particular application based on available data on the
fraction.
1.6.2 Applications in Petroleum Production
(Upstream)
Reservoir, chemical, and mechanical engineers may use the
book in reservoir simulators, design and operation of surface
separators in production fields, and feasibility studies for en-
hanced oil recovery projects, such as gas injection projects.
Another application of the book by reservoir engineers is to
simulate laboratory data on PVT experiments for the reser-
voir fluids, determination of the nature and type of reservoir
fluids, and calculation of the initial amounts of oil and gas in
the reservoir. Reservoir engineers may also use Chapter 9 to
determine the conditions that a solid may form, amount of
solid formation, and method of its prevention during produc-
tion. Practically all chapters of the book should be useful for
reservoir engineers.
1.6.3 Applications in Academia
Although the original goal and aim in writing this book was
to prepare a reference manual for the industry, laboratories,
and research institutions in the area of petroleum, it has been
written in a way such that it can also be used as a textbook
for educational purposes. It can be used as a text for an elec-
tive course for either undergraduate (senior level) or graduate
level. Students from chemical, petroleum, and mechanical en-
gineering fields as well as from chemistry and physics can take
the course and understand the contents of the book. However,
it should not be hard for students from other fields of engi-
neering and science to use this book. The book may also be
used to conduct short courses in the petroleum industry.
1.6.4 Other Applications
There are several other areas in which the book can be used.
One may use this book to determine the quality of crude oils,
petroleum fuels, and products for marketing and government
1. INTRODUCTION 17
organizations that set the standards for such materials. As
an example, the amount of sulfur or aromatic contents of a
fuel can be estimated through minimum laboratory data to
check if they meet the market demand or government regu-
lations for environmental protection. This book can be used
to determine properties of crude oil, its products, and natural
gases that are needed for transportation and storage. Exam-
ples of such properties are density, boiling point, flash and
pour points, sulfur content, vapor pressure, and viscosity.
The book can also be used to determine the properties of
oils for clean-up operations where there is an oil spill on sea-
water. To simulate the fate of an oil spill and the rate of its
disappearance at least the following properties are needed in
order to use appropriate simulators [44, 83-85]:
9 Characterization of petroleum fractions (Chapter 3)
9 Pour point (Chapter 3)
9 Characterization of crude oil (Chapter 4)
9 Solubility parameter (Chapters 4, 6, and 9)
9 Density (Chapters 5 and 7)
9 Vapor pressure (Chapter 7)
9 Viscosity, diffusion coefficient, and surface tension
(Chapter 8)
Accurate prediction of the fate of a crude oil spill depends on
the characterization technique used to estimate the physical
properties. For example, to estimate how much of the ini-
tial oil would be vaporized after a certain time, accurate val-
ues of the diffusion coefficient, vapor pressure, and molecular
weight are needed in addition to an appropriate characteriza-
tion method to split the crude into several pseudocomponents
E833.
1.7 DEFINITION OF UNITS AND
THE CONVERSION FACTORS
An estimated physical property is valuable only if it is ex-
pressed in an appropriate unit. The most advanced process
simulators and the most sophisticated design approaches
fail to perform properly if appropriate units are not used.
This is particularly important for the case of estimation
of physical properties through various correlations or re-
porting the experimental data. Much of the confusion with
reported experimental data arises from ambiguity in their
units. If a density is reported without indicating the tem-
perature at which the density has been measured, this value
has no use. In this part basic units for properties used in
the book are defined and conversion factors between the
most commonly used units are given for each property.
Finally some units specifically used in the petroleum indus-
try are introduced. Interested readers may also find other
information on units from online sources (for example,
http://physics.nist.gov/cuu/contents/index.html).
1.7.1 Importance and Types of U n i t s
The petroleum industry and its research began and grew
mainly in the United States during the last century. The rela-
tions developed in the 1930s, 1940s, and 1950s were mainly
graphical. The best example of such methods is the Winn
nomogram developed in the late 1950s [86]. However, with the
18 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
b i r t h of the c o m p u t e r a n d its expansion, m o r e analytical
m e t h o d s in the f o r m of e q u a t i o n s were developed in the 1960s
a n d m a i n l y in the 1970s a n d 1980s. Nearly all correlations
a n d g r a p h i c a l m e t h o d s that were developed until the early
1980s are in English units. However, starting from the 1980s
m a n y b o o k s a n d h a n d b o o k s a p p e a r e d in the SI units (from Le
Systeme I n t e r n a t i o n a l d'Unites). The general t r e n d is to unify
all engineering b o o k s a n d d o c u m e n t s in SI units to be used
b y the i n t e r n a t i o n a l community. However, m a n y books, re-
ports, h a n d b o o k s , a n d equations a n d figures in various publi-
cations are still in English units. The United States a n d United
K i n g d o m b o t h officially use the English system of units.
Therefore, it is essential t h a t engineers be f a m i l i a r with b o t h
unit systems of English a n d SI. The o t h e r unit system that
is s o m e t i m e s u s e d for s o m e p r o p e r t i e s is the cgs (centimeter,
gram, second) unit, w h i c h is derived f r o m the SI unit.
Since the b o o k is p r e p a r e d for an i n t e r n a t i o n a l audience,
the p r i m a r y unit system u s e d for equations, tables, a n d fig-
ures is the SI; however, it has b e e n tried to p r e s e n t equivalent
of n u m b e r s a n d values of p r o p e r t i e s in b o t h SI a n d English
units. There are s o m e figures t h a t are t a k e n f r o m o t h e r ref-
erences in the literature a n d are in English units a n d t h e y
have been p r e s e n t e d in t h e i r original form. There are s o m e
special units that are c o m m o n l y u s e d to express s o m e spe-
cial properties. F o r example, viscosity is usually expressed in
centipoise (cp), k i n e m a t i c viscosity in centistoke (cSt), den-
sity in g/cm 3, specific gravity (SG) at s t a n d a r d t e m p e r a t u r e
of 60~ o r the GOR in scf/stb. F o r such properties, these pri-
m a r y units have b e e n u s e d t h r o u g h o u t the book, while t h e i r
respective equivalent values in SI are also presented.
1.7.2 F u n d a m e n t a l Units and Prefixes
Generally there are four f u n d a m e n t a l quantities of length (L),
m a s s (M), t i m e (t), a n d t e m p e r a t u r e (T) a n d w h e n t h e i r units
a r e known, units of all o t h e r derived quantities c a n be deter-
mined. In the SI system, units of length, mass, a n d t e m p e r a -
ture are m e t e r (m), k i l o g r a m (kg), a n d Kelvin (K), respectively.
I n English units these d i m e n s i o n s have the units of foot (ft),
p o u n d m a s s Ohm), a n d degrees R a n k i n e (~ respectively. The
u n i t of t i m e in all unit systems is the s e c o n d (s), a l t h o u g h
in English unit, h o u r (h) is also used for the unit of time.
F r o m these units, unit of any o t h e r q u a n t i t y in SI is known.
F o r e x a m p l e the unit of force is SI is kg-m/s 2 w h i c h is called
Newton (N) a n d as a result the unit of p r e s s u r e m u s t be
N/m 2 o r Pascal (Pa). Since 1 Pa is a very small quantity,
larger units such as kPa (1000 Pa) o r m e g a Pascal (MPa)
are c o m m o n l y used. The s t a n d a r d prifixes in SI units are as
follows:
Giga (G) = 109
Mega (M) = 106
Kilo (k) = 103
H e c t o (h) = 102
Deka (da) = 101
Deci (d) = 10 -1
Centi (c) = 10 -2
Milli (m) = 10 -3
Micro (/z) = i 0 6
N a n o (n) = 10 -9
As a n e x a m p l e 1 000 000 Pa can be expressed as 1 MPa. These
prefixes are n o t used in c o n j u n c t i o n with the English units.
However, in the English system of units w h e n v o l u m e t r i c
quantities of gases are p r e s e n t e d in large n u m b e r s , usually
every 1000 units is expressed by one prefix of M. F o r ex-
ample, 2000 scf of gas is expressed as 2 Mscf a n d similarly
2 000 000 scf is w r i t t e n as 2 MMscf. Other s y m b o l s usu-
ally used to express large quantities are b for billion
(1000 million or 109) a n d tr for trillion (one m i l l i o n millions
or 1012).
1.7.3 Units o f Mass
The m a s s is s h o w n b y m a n d its u n i t in SI is kg (kilogram), in
cgs is g (gram), a n d in the English unit system is Ibm (pound-
mass). On m a n y occasions the s u b s c r i p t m is d r o p p e d for lb
w h e n it is referred to mass. In the English u n i t system, units
of o u n c e (oz) a n d grains are also used for m a s s units s m a l l e r
t h a n a p o u n d . F o r larger values of mass, unit of ton is used,
w h i c h is defined in three forms of long, short, a n d metric.
Generally the t e r m t o n is a p p l i e d to the m e t r i c t o n (1000 kg).
The conversion factors are as follows:
1 kg = 1000g = 2.204634 lb --- 35.27392 oz
11b = 0.45359kg = 453.59g = 16oz = 7000 g r a i n
1 g = 0.001 kg = 0.002204634 lb = 15.4324 g r a i n
i ton (metric) = 1000kg = 2204.6341b
1 t o n (short) = 2000 lb = 907.18 kg
I ton (long) = 22401b = 1016kg = 1.12ton (short)
= 1.1016 t o n (metric)
1.7.4 Units o f Length
The u n i t of length in SI is m e t e r (m), in cgs is c e n t i m e t e r
(cm), a n d in English unit system is foot (ft). S m a l l e r values
of length in English system are p r e s e n t e d in i n c h (in.). The
conversion factors are as follows:
1 m = 1 0 0 c m = 10-3 k m = 1000 m m = 106 m i c r o n s (ixm)
= 101~ a n g s t r o m s (A) = 3.28084 ft = 39.37008 in.
= 1.0936yd (yard)
1 ft = 12in. = 0.3048 m = 30.48 c m = 304.8 m m
= 3.048 • 1 0 - 4 k i n = 1 / 3 y d
1 c m = 1 0 - 2 m = 10-5 k m = 1 0 m m = 0.0328084 ft
= 0.393701 in.
l k m = 1 0 0 0 m = 3280.48 ft = 3.93658 x 104in.
1 in. = 2.54 c m = 0.0833333 ft = 0.0254 m = 2.54x 10-5 krn
1 mile = 1609.3m = 1.609km = 5279.8 ft
1.7.5 Units of Time
The unit of t i m e in all m a j o r systems is the s e c o n d (s); how-
ever, for large values of t i m e o t h e r units such as m i n u t e (rain),
h o u r (h), d a y (d), a n d s o m e t i m e s even y e a r (year) are u s e d

appropriately. The conversion factors a m o n g these units are
as follows:
1 year = 365 d = 8 7 6 0 h = 5.256 • 105 m i n = 3.1536 x 107 s
1 d = 2.743973 x 10 3 year = 24 h = 1440 m i n = 8.64 x 104 s
l h = 1.14155 • 10-4year = 4.16667 • 10-2d
= 60 m i n = 3600 s
i m i n = 1.89934 • 10-6year = 6.94444 • 10 -4 d
= 1.66667 x 10-2h = 6 0 s
1 s = 3.17098 x 10 -8 year = 1.157407 • 10 -5 d
= 2.77777 x 10 -4 h = 1.66667 x 10 -2 m i n
1.7.6 U n i t s o f F o r c e
As m e n t i o n e d above, the u n i t of force in the SI system is
N e w t o n (N) an d in the English u n i t system is p o u n d - f o r c e
(lbf). 1 lbf is equivalent to the weight of a mass of 1 lbm at
the sea level w h e r e the acceleration of gravity is 32.174 ft/s 2
(9.807 m/s2). In the cgs system, the unit of force is dyne (dyn).
A n o t h e r unit for the force in the m e t r i c system is kgf, w h i c h
is equivalent to the weight of a mass of 1 kg at the sea level.
The conversion factors are as follows:
1 N = 1 kg. m / s 2 = 105 dyn = 0.2248 lbf = 1.01968 x 10 -1 kgf
1 lbf = 4.4482 N = 0.45359 kgf
1 kgf = 9.807 N = 2.204634 lbf
1 dyn = 10 5 N = 2.248 x 10 -6 lbf
1.7.7 U n i t s o f M o l e s
A not h er unit to present a m o u n t of m a t t e r especially in en-
gineering calculations is mole (mol), w h i c h is defined as the
ratio of mass (m) to m o l e c u l a r weight (M).
m lute and a t m o s p h e r i c pressure. This difference is called gage
(1.6) n = --
M pressure. N o r m a l l y "a" is used to indicate the absolute value
In SI system the unit of m o l e is kmol, w h e r e m in the above
e q u a t i o n is in kg. In the English system, the unit of m o l is
lbmol. In the cgs system, the unit of m o l is gmol, w h i c h is
usually wr i t t en as mol. F o r example, for m e t h a n e (molecular
weight 16.04) i m o l of the gas has mass of 16.04 g. One m o l e
of any substance contains 6.02 x 1023 n u m b e r of molecules
(Avogadro's number). The c o n v e r s i o n factors b e t w e e n vari-
ous units of m o l es are the same as given for the mass in
Section 1.7.3.
1 krnol = l O 0 0 m o l = 2.2046341bmol
1 l b m o l = 0.45359 k m o l = 453.59 m o l
1 tool = 0.001 krnol = 0.002204634 l b m o l
1.7.8 U n i t s o f M o l e c u l a r W e i g h t
M olecu l ar weight or m o l a r mass s h o w n by M is a n u m b e r
that 1 m o l of any substance has equivalent mass of M g. In
the SI system the unit of M is kg/kmol a n d in the English
system the unit is lb/lbmol, while in the cgs system the unit
of M is g/mol. Molecular weight is r e p r e s e n t e d by the s a m e
n u m b e r in all unit systems regardless of the system used. As
an example, m e t h a n e has the m o l e c u l a r w e i g h t of 16 g/mol,
16 lb/lbmol, an d 16 kg/krnol in the unit systems of cgs, SI,
a nd English, respectively. F o r this reason, in m a n y cases the
1. I N T R O D U C T I O N 19
unit for the m o l e c u l a r weight is not m e n t i o n e d ; however, one
m u s t realize that it is not a dimensionless parameter. Most
recent c o m p i l a t i o n s of m o l a r masses are p r o v i d e d by Coplen
[87].
1.7.9 U n i t s o f P r e s s u r e
Pressure is the force exerted by a fluid per unit area; therefore,
in the SI system it has the unit of N / m 2, w h i c h is called Pascal
(Pa), and in the English system has the unit of lbf/ft2 (psf) or
lbf/in. 2 (psi). Other units c o m m o n l y used for the pressure are
the bar (bar) an d standard atmosphere (atm). Pressure m a y
also be expressed in t er m s of m m Hg. In this b o o k units of
MPa, kPa, bar, atm, or psi are c o m m o n l y used for pressure.
The c o n v e r s i o n factors are given as follows:
1 a t m = 1.01325 b ar = 101 3 2 5 P a = 101.325kPa
= 0.101325 MPa = 14.696psi
i a t m = 1.0322 k g f / c m 2 = 760 m m Hg (torr) = 29.921 in. Hg
= 10.333 m H 2 0 (4~
1 b a r = 0.98692 a t m = 1 • 105 Pa = 100kPa
= 0.1 MPa = 14.5038psi
1 Pa = i x 10 - 3 k P a = 1 x 10 - 6 M P a = 9.8692 x 10 - 6 a t m
= 1 • 1 0 - 5 b ar = 1.45037 x 10-4psi
1 psi = 6.804573 • 10 -2 at m = 6.89474 x 1 0 - 2 b a r
= 6.89474 x 10 -3 MPa
1 psf = 144 psi = 9.79858 a t m = 9.92843 b a r = 0.99285 MPa
1 k g f / c m 2 = 0.96784 a t m = 0.98067 b ar = 14.223 psi
The actual pressure of a fluid is the absolute pressure, w h i c h
is m e a s u r e d relative to v a c u u m . However, s o m e pressure
m e a s u r e m e n t devices are calibrated to read zero in the at-
m o s p h e r e an d they s h o w the difference b e t w e e n the abso-
(i.e., psia, bara) an d "g" is used to s h o w the gage pressure
(i.e., psig). However, for absolute pressure very often "a" is
d r o p p e d f r o m the unit (i.e., psi, atm, bar). A n o t h e r unit for
the pressure is v a c u u m pressure that is defined for pressure
b el o w a t m o s p h e r i c pressure. Relations b e t w e e n these units
are as follows:
(1.7) Pgage = Pabs -- Patm
(1.8) Pabs = Patna -- Pvac
Generally gage pressure unit is used to express pressures
above the a t m o s p h e r i c pressures an d v a c u u m pressure unit is
used for pressures b el o w a t m o s p h e r i c and m a y be expressed
in various units (i.e., m m Hg, psi).
1.7.10 U n i t s o f T e m p e r a t u r e
T e m p e r a t u r e (T) is the m o s t i m p o r t a n t p a r a m e t e r affecting
properties of fluids an d it is r ep r esen t ed in Centigrade (~
an d Kelvin (K) in the SI system an d in F a h r e n h e i t (~ and
degrees Ran k i n e (~ in the English unit system. T e m p e r a t u r e
in m o s t eq u at i o n s is in absolute degrees of Kelvin or Rank-
ine. However, acco r d i n g to the definition of Kelvin a nd de-
grees Ran k i n e w h e r e there is a t e m p e r a t u r e difference (AT),
unit of ~ is the s a m e as K an d ~ is the s a m e as ~ These
20 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

t e m p e r a t u r e u n i t s are related t h r o u g h the following relations: this reason, to express a m o u n t of gases i n terms of volume,
n o r m a l l y some SC are defined. The SC i n the metric SI u n i t s
(1.9) T(K) = T(~ + 273.15 are 0~ a n d 1 a t m a n d i n the English system are 60~ a n d
1 atm. U n d e r these conditions m o l a r volume of a n y gas is
(1.10) T(~ = T(~ + 459.67
equivalent to 22.4 L/mol (in SI) a n d 379 scf/lbmol (in English
(1.11) AT(K) = AT(~ units). I n reservoir engineering calculations a n d p e t r o l e u m
i n d u s t r y in general, the SC in the SI units are also set at 60~
(1.12) AT(~ = AT(~
(15.5~ or 289 K) a n d 1 atm. The choice of s t a n d a r d temper-
(1.13) T(~ = 1.8T(K) ature a n d pressure (STP) varies from one source to another.
I n this book w h e n the s t a n d a r d T a n d P are n o t specified the
(1.14) T(~ = 1.8T(~ + 32
STP refers to 289 K a n d I atm, which is equivalent to the STP
in English u n i t system rather t h a n SI system (273 K a n d 1
As a n example, absolute t e m p e r a t u r e of 100 K is equivalent to
atm). However, for liquid systems the volume is less affected
1.8 x 100 or 180~ Therefore, the conversion factors between
by pressure a n d for this reason specification of t e m p e r a t u r e
K a n d ~ are as follows:
alone is sufficient.
1 K = 1.8~ (for absolute t e m p e r a t u r e T a n d the t e m p e r a t u r e
difference, AT) 1.7.12 Units of Volumetric and Mass Flow Rates
1~C -- 1.8 ~F (only for the t e m p e r a t u r e difference, A T) Most processes i n the petroleum i n d u s t r y are c o n t i n u o u s a n d
usually the volume or mass quantities are expressed i n the
1.7.11 Units of Volume, Specific Volume, form of rate defined as volume or mass per u n i t time. One
and Molar V o l u m e - - T h e Standard Conditions particular volumetric flow rate used for liquids i n the English
system is gallon (U.S.) per m i n u t e a n d is k n o w n as GPM.
Volume (V) has the d i m e n s i o n of cubic length (L 3) a n d thus Some of the conversion factors for these quantities are
i n SI has the u n i t of m 3 a n d i n English its u n i t is cubic feet
(cf or ft3). Some units particularly used for liquids in the SI l m 3 / s = 1 • 103 L/s = 1.5851 x 104 GPM
system are liter (L), cm 3 (cc), or milliliter (mL) a n d i n English = 5.4345 x I0 s b b l / d = 1.27133 x 105 fta/h
u n i t s are gallon (in U.S. or Imperial) a n d barrel (bbl). Volume 1 fta/h = 7.86558 x 10-4m3/s = 0.12468GPM
of one u n i t mass of a fluid is called specific volume a n d the = 4.27466 b b l / d
volume of 1 mol of a fluid is called m o l a r volume. Some of 1 GPM = 2.228 x 10 -3 ft3/s = 8.0205 m a / h = 34.285 b b l / d
the conversion factors are as follows.
I b b l / d = 2.9167 x 10 -2 GPM = 1.8401 • 10 -4 ma/s
1 m 3 = 106 cm 3 = 1000 L = 35.315 f t 3 ~- 264.18 gallon (U.S,) = 0.23394 ft3/h
= 35.316 ft 3 = 6.29bbl The conversion factors for the mass rates are as follows:
I ft 3 = 2.8316 x 1 0 - 2 m 3 = 28.316L = 7.4805 gallon(U.S.) 1 kg/s -- 7.93656 x 103 lb/h --- 3.5136 x 107 ton/year
1 bbl = 42 gallon(U.S.) = 158.98 L = 34.973 gallon (Imperial) 1 lb/s = 1.63295 x 103 kg/h = 39.1908ton/d
1 gallon (U.S.) = 0.8327 gallon (Imperial)
= 0.023809 bbl = 3.7853 L The same conversion factors apply to m o l a r rates.

1 mL = i c m 3 = 10-3L = 1 0 - 6 m 3 ---0.061024in. 3
For the m o l a r volumes some of the conversion factors are
given as follows:
1 m 3 / k m o l = 1 L / m o l = 0.001 m 3 / m o l = 1000 c m 3 / m o l
= 16.019 ft3/lbmol
1 ft3/lbmol = 6.24259 x 10 -2 m 3 / k m o l
= 6.24259 x 10 -5 m a / m o l = 62.4259 c m 3 / m o l
1 c m a / m o l = i m L / m o l = 1 L / k m o l = 0.001 m a / k m o l
= 1.6019 x 10 -2 fta/lbmol
It should be noted that the same conversion factors apply
to specific volumes. For example,
1 ft3/lb = 6.24259 • 10 -2 m3/kg = 62.4259cm3/g
Since volume a n d specific or m o l a r volumes depend o n tem-
perature a n d pressure of the system, values of volume in a n y
u n i t system are meaningless if the conditions are n o t spec-
ified. This is particularly i m p o r t a n t for gases in which both
t e m p e r a t u r e a n d pressure strongly influence the volume. For
1.7.13 Units of Density and Molar Density
Density s h o w n by d or p is defined as mass per u n i t volume
a n d it is reciprocal of specific volume. The conversion factors
can be o b t a i n e d from reversing those of specific volume in
Section 1.7.11.
1 kg/m3 = 6.24259 x 10-21b/ft3 = 1 x 10-3 g/cm 3
= 8.3455 x 10 -3 lb/gal
1 lb/ft 3 = 16.019kg/m 3 = 1.6019 • 10-2 g/cm 3
= 0.13368 lb/gal
1 g/cm 3 = 1 kg/L = 103 kg/m 3 = 62.42591b/ft 3
-= 8.3455 lb/gal
1 lb/gal = 1.19825 x 102 kg/m 3 = 7.48031b/ft 3
= 0.119825 g/cm 3
Density m a y also be presented i n terms of n u m b e r of moles
per u n i t volume, which is called molar density a n d is recipro-
cal of m o l a r volume. It can be o b t a i n e d by dividing absolute
density to molecular weight.The conversion factors for m o l a r
density are exactly the same as those for the absolute density
(i.e., I mol/cm 3 = 62.4259 lbmol/ft3). I n practical calculations

the conversion factors may be simplified without major
error in the calculations. For example, 62.4 instead of 62.4259
or 7.48 instead of 7.4803 are used in practical calculations.
In expressing values of densities, similar to specific volumes,
the SC must be specified. Generally densities of liquid hydro-
carbons are reported either in the form of specific gravity at
15.5~ (60~ or the absolute density at 20~ and 1 atm in
g/cm 3.
1.7.14 Units of Specific Gravity
For liquid systems, the specific gravity (SG) is defined as the
ratio of density of a liquid to that of water, and therefore, it is
a dimensionless quantity. However, the temperature at which
specific gravity is reported should be specified. The specific
gravity is also called relative density versus absolute density.
For liquid petroleum fractions and crude oils, densities of
both the oil and water are expressed at the SC of 60~ (15.5~
and 1 atm, and they are usually indicated as SG at 60~176
or simply SG at 60~ Another unit for the specific gravity of
liquid hydrocarbons is defined by the American Petroleum
Institute (API) and is called API degree and is defined in terms
of SG at 60~ (API = 141.5/SG-131.5). For gases, the spe-
cific gravity is defined as the ratio of density of the gas to
that of the air at the SC, which is equivalent to the ratio of
molecular weights. Further discussion on specific gravity, def-
initions, and methods of calculation are given in Chapter 2
(Section 2.1.3).
1.7.15 Units of Composition
Composition is the most important characteristic of homoge-
nous mixtures in which two or more components are uni-
formly mixed in a single phase. Because of the nature of
petroleum fluids, accurate knowledge of composition is im-
portant. Generally composition is expressed as percent-
age (%) or as fraction (percent/100) in terms of weight, mole,
and volume. Density of the components (or pseudocompo-
nents) constituting a mixture is required to convert composi-
tion from weight basis to volume basis or vice versa. Similarly
conversion of composition from mole basis to weight basis
or vice versa requires molecular weight of the constituting
components (or pseudocomponents). Mole, weight, and vol-
ume fractions are shown by Xm, X~, and xv, respectively. Mole,
weight, and volume percentages are shown by mol%, wt%,
and vol%, respectively. Some references use mol/mol, wt/wt,
and vol/vol to express fractional compositions. For normal-
ized compositions, the sum of fractions for all components
in a mixture is 1 (Y~xi = 1) and the sum of all percentages is
100. If the molecular weights of all components in a mixture
are the same, then the mole fraction and weight fraction are
identical. Similarly, if the density (or specific gravity) of all
components is the same, the weight and volume fractions are
identical. The formula to calculate weight fraction from mole
fraction is given as
XmiMi
(1.15) Xwi - - ~N 1 xmiMi
where N is the total number of components, Mi is the molec-
ular weight, and Xwi and Xmi are the weight and mole fractions
of component i, respectively. The conversion from weight to
1. I N T R O D U C T I O N 21
volume fraction can be obtained from the following equation:
Xwi/ SGi
(1.16) xvi -- ~N=I Xwi/SGi
in which x~ is the volume fraction and SGi is the specific grav-
ity of component i. In Eq. (1.16) density (d) can also be used
instead of specific gravity. If mole and weight fractions are
multiplied by 100, then composition is calculated on the per-
centage basis. In a similar way the conversion of composition
from volume to weight and then to mole fraction can be ob-
tained by reversing the above equations. The composition of
a component in a liquid mixture may also be presented by its
molar density, units of which were discussed in Section 1.7.13.
Generally, a solution with solute molarity of 1 has 1 mol of
solute per 1 L of solution (1 mol/L). Through use of both
molecular weight of solute and density of solution one can
obtain weight fraction from molarity. Another unit to express
concentration of a solute in a liquid solution is molality. A so-
lution with molality of 1 has 1 mol of solute per i kg of liquid
solvent.
Another unit for the composition in small quantities is the
ppm (part per million), which is defined as the ratio of unit
weight (or volume) of a component to 106 units of weight or
volume for the whole mixture. Therefore, ppm can be pre-
sented in terms of both volume or weight. Usually in gases
the ppm is presented in terms of volume and in liquids it is
expressed in terms of weight. When ppm is presented in terms
of weight, its relation with wt% is 1 ppm = 10.4 wt%. For ex-
ample, the maximum allowable concentration of H2 S in air
for prolonged exposure is 10 ppm or 0.001 wt%. There is an-
other smaller unit definedas part per billion known as ppb
(1 ppm = 1000 ppb). In the United States a gas is considered
"sweet" if the amount of its H2S content is no more than one
quarter grain per i00 scf of gas. This is almost equivalent to
4 x 10.4 mol fraction [88]. This is in turn equivalent to 4 ppm
on the gas volume basis. Gas composition may also be rep-
resented in terms of partial pressure where sum of all partial
pressures is equivalent to the total pressure.
In general, the composition of gases is presented in volume
or mole fractions, while the liquid composition may be pre-
sented in any form of weight, mole, or volume. For gases at
low pressures (< 1 atm where a gas may be considered an ideal
gas) mole fraction and volume fractions are the same. How-
ever, generally under any conditions, volume and mole frac-
tions are considered the same for gases and vapor mixtures.
For narrow boiling range petroleum fractions with composi-
tions presented in terms of PNA percentages, it is assumed
that densities and molecular weights for all three representa-
tive pseudocompoents are nearly the same. Therefore, with
a good degree of approximation, it is assumed that the PNA
composition in all three unit systems are the same and for
this reason on many occasions the PNA composition is repre-
sented only in terms of percentage (%) or fraction without in-
dicating their weight or volume basis. However, this is not the
case for the crude or reservoir fluid compositions where the
composition is presented in terms of boiling point (or carbon
number) and not in the form of molecular type. The following
example shows conversion of composition from one type to
another for a crude sample.
22 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

TABLE 1.5--Conversion of composition of a crude oil sample from mole to weight and volume percent.
Component tool% Molecularweight (M) Specific
gravity(SG) wt% vol%
C2 0.19 30.07 0.356 0.03 0.06
C3 1.88 44.10 0.508 0.37 0.64
iC4 0.62 58.12 0.563 0.16 0.25
nC4 3.92 58.12 0.584 1.02 1.52
iC5 2.11 72.15 0.625 0.68 0.95
nC5 4.46 72.15 0.631 1.44 1.98
C6 (fraction) 8.59 82.00a 0.690 3.15 3.97
C7+ (fraction) 78.23 266.00 0.895 93.15 90.63
Sum 100.00 100.00 100.00
~This is m o l e c u l a r w e i g h t of C6 h y d r o c a r b o n g r o u p a n d s h o u l d not be m i s t a k e n with M of nC6 w h i c h is 86.2.

Example / . / - - T h e composition of a Middle East crude 1 kW. h = 3600kJ = 3412.2 Btu = 2.655 • 106 ft-lbf
oil is given in Table 1.5 i n terms of mol% with k n o w n
molecular weight a n d specific gravity for each c o m p o n e n t / Energy per u n i t mass is called specific energy that m a y be
p s e u d o c o m p o n e n t . Calculate the c o m p o s i t i o n of the crude i n used to present properties such as specific enthalpy, specific
both wt% a n d vo1%. i n t e r n a l energy, specific heats of reaction, a n d c o m b u s t i o n or
the heating values of fuels. Some of the conversion factors are
Solution--In this table values of molecular weight a n d spe- given below.
cific gravity for pure c o m p o u n d s are o b t a i n e d from Chapter 2
(Table 2.1), while for the C6 group, values are taken from 1 J/g = 103 J/kg = 1 kJ/kg -- 0.42993 Btu/lb
Chapter 4 a n d for the C7 ~ fraction, values are given by the 1 Btu/lb = 2.326 J/g = 0.55556 cal/g
laboratory. Conversion calculations are based o n Eqs. (1.15)
a n d (1.16) o n the percentage basis a n d the results are also The same conversion factors apply to the units of m o l a r en-
given in Table 1.5. I n this calculation it is seen that i n terms ergy such as m o l a r enthalpy.
of wt% a n d vo1%, heavier c o m p o u n d s (i.e., C7+) have higher
values t h a n in terms of mol%. t
1.7.17 Units of Specific Energy per Degrees
1.7.16 Units of Energy and Specific Energy Properties such as heat capacity have the u n i t of specific en-
ergy per degrees. The conversion factors are as follows:
Energy in various forms (i.e., heat, work) has the u n i t of Joule
(1 J -- 1 N- m) i n the SI a n d ft-lbf in the English system. Val- J J
ues of heat are also presented in terms of calorie (in SI) a n d 1~ = 1 x 10 .3 kgOC = 1 = 0.23885 Btu
BTU (British T h e r m a l Unit) i n the English system. There are
two types of joules: absolute joules a n d i n t e r n a t i o n a l joules, cal Btu J
where i Joule (int.) =1.0002 Joule (abs.). In this book only ab- 1~ = 1~ = 4.1867 go--~
solute joules is used a n d it is designated by J. There are also
two types of calories: t h e r m o c h e m i c a l a n d I n t e r n a t i o n a t i o n a l As m e n t i o n e d i n Section 1.7.13, for the difference i n tem-
Steam Tables, where I cal (international steam tables) -- perature (AT), units of ~ a n d K are the same. There-
1.0007 cal (thermochemical) as defined i n the API-TDB [47]. fore, the units of heat capacity m a y also be represented
I n this book cal refers to the i n t e r n a t i o n a l steam tables unless i n terms of specific energy per Kelvin or degrees R a n k i n e
otherwise is specified. I n the cgs system the u n i t of energy is (i.e., 1 ~ = 1 lb.~
Bm = 1 ~ g.~ = 1 ~ ) " The same conversion fac-
~-.~
dyn-cm, which is also called erg. The u n i t of power in the SI tors apply to u n i t s of m o l a r energy per degrees such as m o l a r
system is J/s or watt (W). Therefore, kW.h equivalent to 3600 heat capacity.
kJ is also a u n i t for the energy. The p r o d u c t of pressure a n d Another p a r a m e t e r which has the u n i t of m o l a r energy per
volume (PV) m a y also present the u n i t of energy. Some of the degrees is the universal gas constant (R) used in thermody-
conversion factors for the u n i t s of energy are given as follows: n a m i c relations a n d equations of state. However, the u n i t of
t e m p e r a t u r e for this p a r a m e t e r is the absolute t e m p e r a t u r e
1 J = 1 N. m -- 10 -3 kJ = 107erg -- 0.23885 cal (K or ~ a n d ~ or ~ m a y never be used in this case. Similar
= 9.4783 x 10-4 Btu = 2.778 x 1 0 - 7 k W . h conversion factors as those used for the heat capacity given
1 J = 3.725 x 10 -7 h p . h ---=0.73756 ft.lbf = 9.869 L.atm above also apply to the u n i t s of gas constants in terms of mo-
I cal ( I n t e r n a t i o n a l Tables) -- 3.9683 x 10 -3 Btu = 4.187 J lar energy per absolute degrees.
= 3.088 ft.lbf = 1.1630 x 1 0 - 6 k W . h
Btu cal cal(thermochemical)
1 cal (thermochemical) = 1 cal = 3.9657 x 10 .3 Btu 1 - 1 - 1.0007
lbmol 9~ mol. K mol. K
= 4.184J = 3.086 ft-lbf = 1.1622 x 1 0 - 6 k W . h
J
1Btu = 1055 J = 251.99 cal = 778.16 ft. lbf = 4.1867 x 1 0 3 _
kmol. K
= 2.9307 x 1 0 - 4 k W . h
1 ft-lbf = 1.3558J = 0.32384cal = 1.2851 x 10-3 Btu Numerical values of the gas c o n s t a n t are given i n Sec-
= 3.766 x 1 0 - 7 k W . h tion 1.7.24.

1.7.18 Units of Viscosity and Kinematic Viscosity
Viscosity (absolute viscosity) s h o w n b y / ~ is a p r o p e r t y t h a t
c h a r a c t e r i z e s the fluidity of fluids a n d it has the d i m e n s i o n
of m a s s p e r length p e r t i m e (M/L. t). If the r e l a t i o n b e t w e e n
d i m e n s i o n s of force ( F ) a n d m a s s (M) is u s e d (F = M.L.t-2),
t h e n absolute viscosity finds the d i m e n s i o n of F.t.L -2 w h i c h
is the s a m e as d i m e n s i o n for the p r o d u c t of p r e s s u r e a n d
time. Therefore, in the SI system the unit of viscosity is Pa- s
(N. m-2.s). In the cgs system the unit of viscosity is in g/cm- s
that is called poise (p) a n d its h u n d r e d t h is called centipoise
(cp), w h i c h is equivalent to mi]li-Pa, s ( m P a . s). The conver-
sion factors in various units are given below.
1 cp = 1.02 x 10 -4 kgf. s/m 2 : 1 • 10 -3 Pa- s = 1 m P a . s
= 10-2p = 2.089 x 10 -5 lbf. s/ft 2 = 2.4191b/h- ft
= 3.6 kg/h. m
1 Pa.s = 1 k g / m . s = 1000 cp = 0.67194 lb/ft- s
1 lb/h.ft = 8.634 x 10 -6 lbf. s/ft 2 = 0.4134 cp = 1.488 kg/h. m
i kgf. s/m 2 = 9.804 x 103cp = 9.804 Pa. s = 0.20476 lbf 9s/ft 2
1 lbf. s/ft 2 = 4.788 • 104 cp = 4.884 kgf. s/m 2
The ratio of viscosity to density is k n o w n as kinematic vis-
cosity (v) a n d has the d i m e n s i o n of L/t 2. In the cgs system,
the unit of k i n e m a t i c vsicosity is cm2/s also called stoke (St)
a n d its h u n d r e d t h is centistoke (cSt). The conversion factors
are given below.
1 ft2/h = 2.778 • 10-4 ft2/s --- 0.0929m2/h = 25.81 cSt
1 ft2/s = 9.29 • 104 cSt = 334.5 m2/h
1 cSt = 10 -2 S t - - 10-6m2/s = 1 mmZ/s = 3.875 x 10 -2 ftZ/h
= 1.076 • 10 -s ft2/s
1 m2/s = 104 St = 106 cSt = 3.875 x 104 ft2/h
A n o t h e r u n i t to express k i n e m a t i c viscosity of liquids is
Saybolt universal seconds (SUS), w h i c h is the unit for the
S a y b o l t universal viscosity (ASTM D 88). Definition of viscos-
ity gravity c o n s t a n t (VGC) is b a s e d on SUS unit for the viscos-
ity at two reference t e m p e r a t u r e s of 100 a n d 210~ (37.8 a n d
98.9 ~ The VGC is u s e d in C h a p t e r 3 to e s t i m a t e the com-
p o s i t i o n of heavy p e t r o l e u m fractions. The relation b e t w e e n
SUS a n d cSt is a function of t e m p e r a t u r e a n d it is given in the
API TDB [47]. The analytical relations to convert cSt to SUS
are given b e l o w [47].
SUSeq = 4.6324VT
[1.0 + 0.03264VT]
+ given below.
[(3930.2 + 262.7vr + 23.97v~- + 1.646v 3) x 10 -5]
(1.17)
w h e r e vr is the k i n e m a t i c viscosity at t e m p e r a t u r e T in cSt.
The SUSeq calculated from this relation is converted to the
S U S r at the d e s i r e d t e m p e r a t u r e of T t h r o u g h the following
relation.
(1.18) SUST = [1 + 1.098 x 10-4(T - 311)]SUSeq
w h e r e T is the t e m p e r a t u r e in kelvin (K). F o r conversion of
cST to SUS at the reference t e m p e r a t u r e of 311 K (100~
only Eq. (1.17) is needed. E q u a t i o n (1.18) is the c o r r e c t i o n
t e r m for t e m p e r a t u r e s o t h e r t h a n 100~ F o r k i n e m a t i c vis-
cosities g r e a t e r t h a n 70 cSt, Eqs. (1.17) a n d (1.18) c a n be
1. I N T R O D U C T I O N 23
simplified to the following form at the t e m p e r a t u r e s of 311
(100~ a n d 372 K, (210~ respectively [1].
(1.19) SUS100F = 4.632v100F vl00F > 7 5 c S t
(1.20) SUS210F = 4.664VZ10F V210F> 75 cSt
w h e r e Vl00F is the k i n e m a t i c viscosity at 100~ (311 K) in cSt.
As a n example, a p e t r o l e u m fraction with k i n e m a t i c viscosity
of 5 cSt at 311 K has a n equivalent S a y b o h Universal Viscosity
of 42.4 SUS as calculated f r o m Eq. (1.17).
A n o t h e r unit for the viscosity is S F S (Saybolt foural sec-
onds) expressed for S a y b o l t foural viscosity, w h i c h is m e a -
s u r e d in a way s i m i l a r to S a y b o l t universal viscosity b u t
m e a s u r e d b y a larger orifice (ASTM D 88). The conversion
f r o m cSt to S F S is expressed t h r o u g h the following equations
at two reference t e m p e r a t u r e s of 122~ (323 K) a n d 210~
(372 K) [47].
13924
(1.21) SFS122F = 0.4717Vm22F +
U 1F2 22 -- 72.59VI22F + 6816
5610
(1.22) SFS210F = 0.4792v2mF + v2210F+ 2130
F o r conversion of Saybolt foural viscosity (SFS) to k i n e m a t i c
viscosity (cSt.), the above equations should be u s e d in reverse
o r to use t a b u l a t e d values given b y API-TDB [47]. As an exam-
ple, an oil w i t h Saybolt foural viscosity of 450 S F S at 210~
has a k i n e m a t i c viscosity of 940 cSt. Generally, viscosity of
highly viscous oils is p r e s e n t e d by SUS or S F S units.
1.7.19 Units of Thermal Conductivity
T h e r m a l conductivity (k) as discussed in C h a p t e r 8 r e p r e s e n t s
a m o u n t of h e a t passing t h r o u g h a u n i t a r e a of a m e d i u m for
one unit of t e m p e r a t u r e g r a d i e n t ( t e m p e r a t u r e difference p e r
unit length). Therefore, it has the d i m e n s i o n of energy p e r
t i m e p e r a r e a p e r t e m p e r a t u r e gradient. In the SI units it is
expressed in J/s- m . K. Since t h e r m a l conductivity is defined
b a s e d on a t e m p e r a t u r e difference (AT), the u n i t of ~ m a y
also be used i n s t e a d of K. Because J/s is defined as w a t t (W),
the u n i t of t h e r m a l conductivity in the SI system is u s u a l l y
w r i t t e n as W/re. K. I n the English system, the unit of ther-
m a l conductivity is Btu a n d in s o m e references is w r i t t e n
as ~9' h .o~ t/" w h i c h is the. ratio of h e a t flux to the. t e m p e r. a t u r e
gradient. The conversion factors b e t w e e n various u m t s are
1 W/re. K (J/s. m.~ = 0.5778 Btu/fl. h.~
= 1.605 • 10 -4 Btu/ft. s.~
= 0.8593 kcal/h- m-~
1 Btu/ft-h-~ = 1.7307 W / m . K
1 cal/cm 9s.~ = 242.07 Btu/ft 9h-~ = 418.95 W / m . K
1.7.20 Units of Diffusion Coefficients
Diffusion coefficient o r diffusivity r e p r e s e n t s the a m o u n t of
m a s s diffused in a m e d i u m p e r unit a r e a p e r unit t i m e p e r
unit c o n c e n t r a t i o n gradient. As s h o w n in C h a p t e r 8, it has
the s a m e d i m e n s i o n as the k i n e m a t i c viscosity, w h i c h is
24 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
squared length per time (L2]t). Usually it is expressed i n
cm2/s.
i cm2/s = 10 -4 m2/s : 9.29 x 10 -6 f t 2 / s : 3.3445 x 10 -4 f t 2 / h
1.7.21 Units o f Surface Tension
Surface t e n s i o n or interfacial t e n s i o n (a) as described i n Sec-
tion 8.6 (Chapter 8) has the u n i t of energy (work) per u n i t area
a n d the SI u n i t of surface t e n s i o n is J/m 2 = N/re. Since N/m
is a large u n i t the values of surface t e n s i o n are expressed in
milli-N/m (mN/m) which is the same as the cgs u n i t of surface
t e n s i o n (dyn/cm). The conversion factors for this property are
as follows:
1 dyn/cm = 1 erg/cm 2 = 10 -3 J/m 2 = 1 mJ/m 2
= 10 -3 N/m -- I m N / m
1.7.22 Units o f Solubility Parameter
Prediction of solubility p a r a m e t e r (~) for p e t r o l e u m fractions
a n d crude oil is discussed in Chapters 4 a n d 10 a n d it has
the u n i t of (energy/volume)~ The traditional u n i t of ~ is i n
(cal/cm3) ~ Another form of the u n i t for the solubility pa-
r a m e t e r is (pressure) ~ Some conversion factors are given
below.
1 (calth/cm3) ~ = 2.0455 (J/cm3) ~ = 2.0455 (MPa) ~
= 2.0455 x 103 (j/m3) ~
= 2.0455 x 103 (Pa) ~ = 10.6004 (Btu/ft3) ~
= 31.6228 (kcalth/m 3)~
6.4259 (atm) ~ = 2.05283 (ft-lbf/ft3) ~
1 (MPa) ~ = 0.4889 (calth/cm3) ~ = l(J/cm3) ~ = 103 (Pa) ~
Values of surface t e n s i o n in the literature are usually ex-
pressed i n (cal/cm3) ~ where cal represents t h e r m o c h e m i c a l
u n i t of calories.
1.7.23 Units o f Gas-to-Oil Ratio
Gas-to-oil ratio is a n i m p o r t a n t p a r a m e t e r i n d e t e r m i n i n g the
type of a reservoir fluid a n d i n setting the o p t i m u m operating
conditions in the surface separators at the p r o d u c t i o n field
(Chapter 9, Section 9.2.1). I n some references such as the
API-TDB [47], this p a r a m e t e r is called gas-to-liquid ratio a n d
is s h o w n by GLR. GOR represents the ratio of volume of gas to
the volume of liquid oil from a separator u n d e r the S C of 289 K
a n d 101.3 kPa (60~ a n d 14.7 psia) for both the gas a n d liquid.
Units of volume were discussed i n Section i. 7.13. Three types
of u n i t s are c o m m o n l y used: the oilfield, the metric, a n d the
English units.
9 Oilfield units: s t a n d a r d cubic feet (scf) is used for the volume
of gas, a n d stock tank barrels (stb) is used for the v o l u m e of
oil. Therefore, GOR has the u n i t of scf/stb.
9 Metric units: s t a n d a r d cubic meters (sm 3) is used for the
gas, a n d stock t a n k cubic meters (stm 3) u n i t is used for the
oil. The volume of liquid oil p r o d u c e d is usually presented
u n d e r the stock t a n k conditions, which are 60~ (15.5~
a n d 1 atm. Therefore, GOR u n i t in this system is sma/stm 3.
9 English unit: scf is used for the gas, a n d sock tank cubic feet
(stft 3) is used for the liquid volume. Thus the GOR has the
u n i t s of scf/stft 3. This u n i t is exactly the same as sm3/stm 3
i n the SI unit.
The conversion factors b e t w e e n these three u n i t s for the GOR
(GLR) are given as follows:
1 scf/stb = 0.1781 scf/stft 3 -- 0.1781 sm3/stm 3
I sm3/stm 3 : I scf/stft 3 = 5.615 scf/stb
1.7.24 Values o f Universal Constants
I. 7.24.1 Gas Constant
The universal gas c o n s t a n t s h o w n by R is used in equations
of state a n d t h e r m o d y n a m i c relations in Chapters 5, 6, 8, a n d
10. It has the u n i t of energy per mole per absolute degrees. As
discussed i n Section 1.7.17, its d i m e n s i o n is similar to that of
m o l a r heat capacity. The value of R i n the SI u n i t is 8314 J/
kmol- K. The energy d i m e n s i o n m a y also be expressed as the
p r o d u c t of pressure a n d volume (PV), which is useful for ap-
plication in the equations of state. Value of R in terms of en-
ergy u n i t is more useful i n the calculation of t h e r m o d y n a m i c
properties such as heat capacity or enthalpy. Values of this
p a r a m e t e r i n several other u n i t s are given as follows.
R -- 8.314 J/mol 9K = 8314 J/kmol 9K = 8.314 mapa/mol - K
= 83.14 cm3bar/mol 9K
= 82.06 c m 3.atm/mO1 9K = 1.987 calth/mol - K
= 1.986 cal/mol 9K = 1.986 Btu/lbmol 9R
-- 0.7302 ft3-atm/lbmOl' R = 10.73 ft 3.psia/lbmol 9R
= 1545 ft. lbf/lbmol 9R
1.7.24.2 Other Numerical Constants
The Avogadro n u m b e r is the n u m b e r of molecules i n 1 mol of
a substance.
NA = Avogadro n u m b e r -- 6.022 x 1023 mo1-1
For example 1 mol of m e t h a n e (16 g) consists of 6.022• 1023
molecules. Other constants are
B o l t z m a n c o n s t a n t = kB -- R/NA = 1.381 x 10 -23 J/K.
Planck c o n s t a n t = h = 6.626 • 10 -34 J. s.
Speed of light in v a c u u m = c = 2.998 x 108 m/s.
N u m e r i c a l constants
~r = 3.14159265
e = 2.718281 828
lnx = lOgl0 x/logl0e = 2.30258509 log10x.
1.7.25 Special Units for the Rates and A m o u n t s
o f Off and Gas
A m o u n t s of oil a n d gas are usually expressed in volumetric
quantities. I n the p e t r o l e u m i n d u s t r y the c o m m o n u n i t for
volume of oil is barrel (bbl) a n d for the gas is s t a n d a r d cubic
feet (scf) both at the conditions of 60~ (15.5~ a n d 1 atm.
The p r o d u c t i o n rate for the crude is expressed i n bbl/d a n d
for the gas i n scf/d.
I n some cases, a m o u n t of crude oil is expressed i n the met-
ric ton. Conversion from volume to weight or vice versa re-
quires density or specific gravity (API) of the oil. For a light
 1. INTRODUCTION 25

Saudi Arabian crude of 35.5 API (SG = 0.847), the follow- factors with ratios of unity as follows:
ing conversion factors apply between weight and volume of
crudes and the rates: 802~ol =(802~ol) x m
X ~l 6 g ~1 0 ~ J x

1 ton ~ 7.33 bbl = 308gallon (U.S.) 1 bbl ~- 0.136ton /802 x 1000\

- ~ ~-x--4.-i-~) [cal/g] = l1971.58cal/g
1 bbl/d ~ 50 ton/year
The conversion to the English unit is performed in a similar
For a Middle East crude of API 30, 1 ton - 7.19 bbl (1 bbl
way:
0.139 ton).
Another way of expressing quantities of various sources of 453.6 g Btu
11971.58cal/g=(l1971.58cal/g) x ~ x 252cal
energy is through their heating values. For example, by burn-
ing 1 x 106 tons of a crude oil, the same a m o u n t of energy (11971.58 x 4 5 3 . 6 )
can be produced that is p r o d u c e d through burning 1.5 x 109 = ~ [Btu/lb]
tons of coal. Of course this value very m u c h depends on the
type of crude and the coal. Therefore, such evaluations and = 21549.2 Btu/lb
comparisons are approximate. In summary, 1 million tons of
In the above calculations all the ratio of terms inside the II
a typical crude oil is equivalent to other forms of energy:
sign have values of unity.
1 x 106tons of crude oil ~ 1.111 x 1 0 9 s m 3 (39.2 x 109scf)
Example 1 . 3 - - T h e r m a l conductivity of a kerosene sample at
of natural gas
60~ is 0.07 Btu/h - ft-~ What is the value of thermal conduc-
1.5 x 109tons o f c o a l
tivity in m W / m K from the following procedures:
---- 12 x 109 kW. h of electricity
1. Use of appropriate conversion factor in Section 1.7.19.
The --- sign indicates the approximate values, as they depend 2. Direct calculation with use of conversion factors for fun-
on the type of oil or gas. For a typical crude, the heating value damental dimensions.
is approximately 10 500 cal/g (18 900 Btu/lb) and for the nat-
ural gas is about 1000 Btu/scf (37.235 x 103 kJ/sm3). Approx- Solution--
imately 1 million tons of a typical crude oil can produce an
1. In Section 1.7.19 the conversion factor between SI and
energy equivalent to 4 x 109 kW- h of electricity through a typ-
English units is given as:
ical power plant. In 1987 the total nuclear energy produced
1 W/mK = 0.5778 Btu/ft. h.~ With the knowledge that
in the world was equivalent to 404 x 106 tons of crude oil
W = 1000 rnW, the conversion is carried as:
based on the energy produced [5]. In the same year the to-
tal hydroelectric energy was equivalent to 523.9 x 106 tons of 0.07Btu/h.ft.~ \.(007 h.ft.oF]xBtU
1~ 100OmW
crude oil. In 1987 the total coal reserves in the world were
estimated at 1026 x 109 tons, while the total oil reserves were
about 122 x 109 tons. However, from the energy point of view W/mK ~
x 0 . 5 7 7 8 B t u / h - ft - =121.1 m W / m K
the total coal reserves are equivalent to only 0.68 x 109 tons
of crude oil. The subject of heating values will be discussed 2. The conversion can be carried out without use of the con-
further in Chapter 7 (see Section 7.4.4). version tables if a practicing engineer is familiar with the
Unit conversion is an important art in engineering calcu- basic definitions and conversion factors. These are 1 W =
lations and as was stated before with the knowledge of the 1 J/s, 1 W = 1000 mW, 1 cal = 4.187 J, I Btu = 251.99 cal,
definition of some basic units for only a few fundamental 1 h = 3600 s, 1 ft = 0.3048 m, 1 K = I~ = 1.8~ (for the
quantities (energy, length, mass, time, and temperature), the temperature difference). It should be noted that thermal
unit for every other property can be obtained. The basic idea conductivity is defined based on temperature difference.
in the unit conversion is that a value of a parameter remains
the same when it is multiplied by a factor of unity in a way 0.07 Btu/h - ft. ~
that the initial units are eliminated and the desired units are [ 0 0 7 [ Btu 1 251.99cal 4.187J 3 6 @ 0s I
kept. The following examples demonstrate how a unit can be =[ " ]h.ft.~ x Btu x cal x
converted to another unit system without the use of tabulated
conversion factors. x ~/s x 1 0 0 ~ W x 0.3~48m x 1'~ x

Example 1.2--The molar heating value of m e t h a n e is 802 kJ/ [0.07 x 251.99 x 4.187 x 1000 x 1.8 m_~.~
mol. Calculate the heating value of methane in the units of
cal/g and Btu/lb. The molecular weight of methane is 16.0,
= 121.18 m W / m K

Solution--In this calculation a practicing engineer has to re-
m e m b e r the following basic conversion factors: 1 lb = 453.6 g,
1 cal = 4.187 J, and 1 Btu = 252 cal. The value of molecular
weight indicates that 1 tool = 16 g. In the conversion process
the initial unit is multiplied by a series of k n o w n conversion
Examples 1.2 and 1.3 show that with the knowledge of only
very few conversion factors and basic definitions of funda-
mental units, one can obtain the conversion factor between
any two unit systems for any property without use of a refer-
ence conversion table.
26 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

1.8 PROBLEMS Europe, Africa, Middle East, Eastern Europe and

Former Soviet Union, and Asia Pacific)?
1.1. State one theory for the formation of petroleum and give b. Where is the location of the biggest refinery in the
names of the hydrocarbon groups in a crude oil. What world and what is its capacity in bbl/d?
are the most important heteroatoms and their concen- c. What is the history of the rate of production of gaso-
tration level in a crude oil? line, distillate, and residual from refineries in the
1.2. The following compounds are generally found in the world and the United States for the last decade?
analysis of a crude oil: ethane, propane, isobutane, 1.9. Characteristics of three reservoir fluids are given below.
n-butane, isopentane, n-pentane, 2,2-dimethylbutane, For each case determine the type of the reservoir fluid
cyclopentane, cyclohexane, n-hexane, 2-methylpentane, using the rule of thumb.
3-methylpentane, benzene, methylcyclopentane, 1,1- a. GOR = 20 scf/stb
dimethylcyclopentane, and hydrocarbons from C7 and b. GOR = 150 000 scf/stb
heavier grouped as C7+. c. CH4 mol% -- 70, API gravity of STO = 40
a. For each compound, draw the chemical structure and 1.10. GOR of a reservoir fluid is 800 scf/stb. Assume the molec-
give the formula. Also indicate the name of hydrocar- ular weight of the stock tank oil is 260 and its specific
bon group that each compound belongs to. gravity is 0.87.
b. From the above list give the compounds that possibly a. Calculate the GOR in sma/stm a and the mole fraction
exist in a gasoline fraction. of gases in the fluid.
1.3. Give the names of n-C20, n-C3o, n-C40, and three isomers b. Derive a general mathematical relation to calculate
of n-heptane according to the IUPAC system. GOR from mole fraction of dissolved gas (XA)through
1.4. List the 10 most important physical properties of crude STO gravity (SG) and oil molecular weight (M).
and its products that are required in both the design and Calculate XAusing the developed relation.
operation of an atmospheric distillation column. 1.11.The total LPG production in 1995 was 160 million
1.5. What thermodynamic and physical properties of gas tons/year. If the specific gravity of the liquid is assumed
and/or liquid fluids are required for the following two to be 0.55, what is the production rate in bbl/d?
cases? 1.12. A C7+ fraction of a crude oil has the following composi-
a. Design and operation of an absorption column with tion in wt%. The molecular weight and specific gravity of
chemical reaction [40, 89]. each pseudocomponent are also given below. Calculate
b. Reservoir simulation [37]. the composition of crude in terms of vol% and mol%.
1.6. What is the characterization of petroleum fractions,
crude oils, and reservoir fluids? Explain their differ- Pseudocomponent wt% M SG
ences. C7+ (1) 17.3 110 0.750
1.7. Give the names of the following compounds according C7+ (2) 23.6 168 0.810
to the IUPAC system. C7+ (3) 31.8 263 0.862
C7+ (4) 16.0 402 0.903
C7+ (5) 11.3 608 0.949
Total C7+ 100
a
1.13. It is assumed that a practicing engineer remembers the
following fundamental unit conversion factors without
a reference.
1 ft = 0.3048 m -- 12 in.
1 atm = 101.3 kPa = 14.7 psi
1 K = 1.8~
1 Btu = 252 cal
1 cal = 4 . 1 8 J
1 kg = 2.2 lb
g = 9.8 m/s 2
1 lbmol = 379 scf
Molecular weight of methane -- 16 g/tool
(b) (c) (d) Calculate the following conversion factors using the
above fundamental units.
a. The value of gas constant is 1.987 cal/mol 9K. What is
e. CH2-~-~CH-- C H ~ C H - - CHz---CH3 its value in psi. ft3/lbmol 9R?
b. Pressure of 5000 psig to atm
c. 1 kgf/cm2 to kPa
1.8. From an appropriate reference find the following data d. 1 Btu/lb.~ to J/kg. K
in recent years. e. 1 Btu/lbmol to cal/g
a. What is the distribution of refineries in different parts f. 1000 scf of methane gas to lbmol
of the world (North America, South America, Western g. 1 MMM scf of methane to kg

h. 1 cp to lb/ft.h
i. 1 P a . s to cp
j. 1 g/cm 3 to lb/fta
1.14. A crude oil has API gravity of 24. W h a t is its density i n
g/cm 3, lb/ft 3, kg/L, kg/m3 ?
1.15. Convert the following u n i t s for the viscosity.
a. Crude viscosity of 45 SUS (or SSU) at 60~ (140~ to
cSt.
b. Viscosity of 50 SFS at 99~ (210~ to cSt.
c. Viscosity of 100 cp at 38~ (100~ to SUS
d. Viscosity of 10 cp at 99~ (210~ to SUS
1.16. For each t o n of a typical crude oil give the equivalent
estimates in the following terms:
a. Volume of crude i n bbl.
b. Tons of equivalent coal.
c. S t a n d a r d cubic feet (scf) a n d sm 3 of n a t u r a l gas.
1.17. I n terms of equivalent energy values, compare exist-
ing reserves for three m a j o r fossil types a n d n o n r e n e w -
able sources of energy: oil, n a t u r a l gas, a n d coal by
calculating
a. the ratio of existing world total gas reserves to the
world total oil reserves.
b. the ratio of existing world total coal reserves to the
world total oil reserves.
c. the p e r c e n t share of a m o u n t of each energy source in
total reserves of all three sources.
REFERENCES
[1] Speight, J. G, The Chemistry and Technology of Petroleum, 3rd
ed., Marcel Dekker, New York, 1998.
[2] Jahn, F., Cook, M., and Graham, M., Hydrocarbon Exploration
and Production, Developments in Petroleum Science 46,
Elsevier, Amsterdam, 1998.
[3] Oil and Gas Journal Data Book, 2000 edition, PennWell
Publishing, Tulsa, OK, 2000.
[4] International Petroleum Encyclopedia, Volume 32, PennWell
Publishing, Tulsa, OK, 1999.
[5] BP Statistical Review of World Energy, British Petroleum,
United Kingdom, June 1988.
[6] Basic Petroleum Data Book, Petroleum Industry Statistics,
Volume XVIII, Number i, American Petroleum Institute (API),
Washington, DC, 1988.
[7] The International Union of Pure and Applied Chemistry
(IUPAC), 2003, Research Triangle Park, NC, URL:
http://www.iupac.org.
[8] Nelson, W. L., Petroleum Refinery Engineering, McGraw-Hill,
New York, 1958.
[9] Manning, F. S. and Thompson, R. R., Oilfield Processing,
PennWell Publishing, Tulsa, OK, 1995.
[10] Altagelt, K. H. and Boduszynski, M. M., Composition and
Analysis of Heavy Petroleum Fractions, Marcel Dekker, New
York, 1994.
[11] McCain, Jr., W. D., The Properties of Petroleum Fluids, 2nd ed.,
PennWell Publishing, Tulsa, OK, 1990.
[12] Speight, J. G., Gas Processing--Environmental Aspects and
Methods, Butterworth Heinemann, Oxford, England, 1993.
[13] Pedersen, K. S., Fredenslund, Aa., and Thomassen, E, Properties
of Oils and Natural Gases, Gulf Publishing, Houston, TX, 1989.
[14] Arnold, K. and Stewart, M., Surface Production Operations,
Design of Oil-Handling Systems and Facilities, Volume 1, 2nd
ed., Gulf Publishing, Houston, TX, 1998.
1. I N T R O D U C T I O N 27
[15] Favennec, J. E, Petroleum Refining, Volume 5: Refinery
Operation and Management, Editions Technip, Paris,
1998.
[16] Mansoori, G. A., "Modeling of Asphaltene and Other Heavy
Organic Depositions," Journal of Petroleum Science and
Engineering, Vol. 17, 1997, pp. 101-111.
[17] Goual, L. and Firoozabadi, A., "Measuring Asphaltenes and
Resins, and Dipole Moment in Petroleum Fluids," American
Institute of Chemical Engineers Journal, Vol. 48, No. 11, 2002,
pp. 2646-2662.
[18] Wauquier, J.-E, Petroleum Refining, Volume 1: Crude Oil,
Petroleum Products, Process Flowsheets, Editions Technip,
Paris, 1995.
[19] Gary, J. H. and Handwerk, G. E., Petroleum Refining--
Technology and Economics, 3rd ed., Marcel & Dekker, New
York, 1994.
[20] Beck, R. J., "Demand Growth to Continue for Oil, Resume for
Gas this Year in the U.S.," Oil and Gas Journal, January 26,
1998, p. 57.
[21] Totten, G. E., Shah, R. J., and Westbrook, S. R., Fuels and
Lubricants Handbook: Technology, Properties, Performance, and
Testing, Eds., ASTM MNL 37, ASTM International, West
Conshohocken, PA, 2003, pp. 1075.
[22] ASTM, Annual Book of Standards, Section Five, Petroleum
Products, Lubricants, and Fossil Fuels (in Five Volumes),
ASTM International, West Conshohocken, PA, 2002.
[23] Riazi, Mi R. and Daubert, T. E., "Simplify Property Predictions,"
Hydrocarbon Processing, Vol. 59, No. 3, 1980, pp. 115-116.
[24] Riazi, M. R. and Daubert, T. E., "Characterization Parameters
for Petroleum Fractions," Industrial and Engineering Chemistry
Research, Vol. 26, 1987, pp. 755-759.
[25] Riazi, M. R. and A1-Otaibi, G. N., "Estimation of Viscosity of
Petroleum Fractions," Fuel, Vol. 80, 2001, pp. 27-32.
[26] Riazi, M. R. and Daubert, T. E., "Molecular Weight of Heavy
Fractions from Viscosity," Oil and Gas Journal, Vol. 58, No. 52,
1987, pp. 110-113.
[27] Riazi, M. R. and Daubert, T. E., "Prediction of the Composition
of Petroleum Fractions," Industrial and Engineering Chemistry,
Process Design and Development, Vol. 19, No. 2, 1980,
pp. 289-294.
[28] Riazi, M. R. and Daubert, T. E., "Prediction of Molecular Type
Analysis of Petroleum Fractions and Coal Liquids," Industrial
and Engineering Chemistry, Process Design and Development,
Vol. 25, No. 4, 1986, pp. 1009-1015.
[29] Riazi, M. R., Nasimi, N., and Roomi, Y., "Estimating Sulfur
Content of Petroleum Products and Crude Oils," Industrial and
Engineering Chemistry Research, VoL 38, No. 11, 1999,
pp. 4507-4512.
[30] Riazi, M. R., A1-Enezi, G., and Soleimani, S., "Estimation of
Transport Properties of Liquids," Chemical Engineering
Communications, Vol. 176, 1999, pp. 175-193.
[31] Riazi, M. R. and Roomi, Y., "Use of the Refractive Index in the
Estimation of Thermophysical Properties of Hydrocarbons and
Their Mixtures," Industrial and Engineering Chemistry
Research, Vol. 40, No. 8, 2001, pp. 1975-1984.
[32] Riazi, M. R., Prediction of Thermophysical Properties of
Petroleum Fractions, Doctoral Dissertation, Department of
Chemical Engineering, Pennsylvania State University,
University Park, PA, 1979.
[33] Riazi, M. R. and Al-Sahhaf, T., "Physical Properties of
n-Alkanes and n-Alkyl Hydrocarbons: Application to Petroleum
Mixtures," Industrial and Engineering Chemistry Research,
Vol. 34, 1995, pp. 4145-4148.
[34] Daubert, T. E., "Property Predictions," Hydrocarbon Processing,
Vol. 59, No. 3, 1980, pp. 107-112.
[35] Riazi, M. R. and Roomi, Y. A., "Compositional Analysis of
Petroleum Fractions," Preprints of Division of Petroleum
28 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
Chemistry Symposia, American Chemical Society (ACS),
Vol. 40, No. 8, 2001, pp. 1975-1980.
[36] Poling, B. E., Prausnitz, J. M., and O'Connell, J. E, Properties of
Gases and Liquids, 5th ed., McGraw-Hill, New York, 2000.
[37] Mattax, C. C. and Dalton, R. L., Reservoir Simulation, SPE
Nomograph 13, Society of Petroleum Engineers, Richardson,
TX, 1990.
[38] Riazi, M. R. and Faghri, A., "Effect of the Interfacial Drag on
Gas Absorption with Chemical Reaction in a Vertical Tube,"
American Institute of Chemical Engineers Journal,
Vol. 32, No. 4, 1986, pp. 696-699.
[39] Dabir, B., Riazi, M. R., and Davoudirad, H. R., "Modeling of
Falling Film Reactors," Chemical Engineering Science, Vol. 51,
No. 11, 1996, pp. 2553-2558.
[40] Riazi, M. R., Whitson, C. H., and da Silva, E, "Modelling of
Diffusional Mass Transfer in Naturally Fractured Reservoirs,"
Journal of Petroleum Science and Engineering, Vol. 10, 1994,
pp. 239-253.
[41] Riazi, M, R. and Whitson, C. H., "Estimating Diffusion
Coefficients of Dense Fluids," Industrial and Engineering
Chemistry Research, Vol. 32, No. 12, 1993, pp. 3081-3088.
[42] Riazi, M. R., "A New Method for Experimental Measurement of
Diffusion Coefficient in Reservoir Fluids," Journal of Petroleum
Science and Engineering, Vol. 14, 1996, pp. 235-250.
[43] Kister, H. Z., Distillation Operations, McGraw-Hill, New York,
1990.
[44] Villoria, C. M., Anselmi, A. E., Intevep, S. A., and Garcia, E R.,
"An Oil Spill Fate Model," SPE23371, Society of Petroleum
Engineers, 1991, pp. 445-454.
[45] Whitson, C. H. and Brule, M. R., Phase Behavior, Monograph
Volume 20, Society of Petroleum Engineers, Richardson, TX,
2000.
[46] Firoozabadi, A., Thermodynamics of Hydrocarbon Reservoirs,
McGraw-Hill, New York, 1999.
[47] Daubert, T. E. and Danner, R. E, Eds., API Technical Data
Book--Petroleum Refining, 6th ed., American Petroleum
Institute (API), Washington, DC, 1997.
[48] AIChE DIPPR| Design Institute for Physical
Property Data (DIPPR), EPCON International, Houston, TX,
1996.
[49] Rowley, R. L., Wilding, W. V., Oscarson, J. L., Zundel, N. A.,
Marshall, T. L., Daubert, T. E., and Danner, R. P., DIPPR Data
Compilation of Pure Compound Properties, Design Institute for
Physical Properties (DIPPR), Taylor & Francis, New York, 2002
(http://dippr.byu.edu).
[50] Thermodynamic Research Center, National Institute of
Standards and Technology (NIST), Boulder, CO, URL:
http://www.nist.gov/.
[51] Riazi, M. R. and Daubert, T. E., "Analytical Correlations
Interconvert Distillation Curve Types," Oil and Gas Journal,
August 25, 1986, pp. 50-57.
[52] Riazi, M. R. and Daubert, T. E., "Improved Characterization of
Wide Boiling Range Undefined Petroleum Fractions," Industrial
and Engineering Chemistry Research, Vol. 26, 1987, pp. 629-632.
[53] Riazi, M. R. and Daubert, T. E., "Predicting Flash Points and
Pour Points of Petroleum Fractions," Hydrocarbon Processing,
September 1987, pp. 81-84.
[54] Riazi, M. R., "A Distribution Model for C7+ Fractions
Characterization of Petroleum Fluids," Industrial and
Engineering Chemistry Research, Vol. 36, 1997, pp. 4299-4307.
[55] Riazi, M. R., "Distribution Model for Properties of
Hydrocarbon-Plus Fractions," Industrial and Engineering
Chemistry Research, Vol. 28, 1989, pp. 1731-1735.
[56] Riazi, M. R. and Al-Sahhaf, T. A., "Physical Properties of Heavy
Petroleum Fractions and Crude Oils," Fluid Phase Equilibria,
Vol. 117, 1996, pp. 217-224.
[57] Watson, IC M., Nelson, E. E, and Murphy, G. B.,
"Characterization of Petroleum Fractions," Industrial and
Engineering Chemistry, Vol. 27, 1935, pp. 1460-1464.
[58] Brule, M. R., Kumar, K. H., and Watanasiri, S.,
"Characterization Methods Improve Phase Behavior
Predictions," Oil & Gas Journal, February 11, 1985, pp. 87-93.
[59] Riazi, M. R., A1-Sahhaf, T. A., and Al-Shammari, M. A.,
"A Generalized Method for Estimation of Critical Constants,"
Fluid Phase Equilibria, Vol. 147, 1998, pp. 1-6.
[60] Dohrn, R. and Pfohl, O. "Thermophysical
Properties--Industrial Directions," Paper presented at the
Ninth International Conference on Properties and Phase
Equilibria for Product and Process Design (PPEPPD 2001),
Kurashiki, Japan, May 20-25, 2001.
[61] McCabe, W. L., Smith, J. C., and Harriot, P., Unit Operations of
Chemical Engineering, 4th ed., McGraw-Hill, New York, 1985.
[62] Peridis, S., Magoulas, K., and Tassios, D., "Sensitivity of
Distillation Column Design to Uncertainties in Vapor-Liquid
Equilibrium Information," Separation Science and Technology,
Vol. 28, No. 9, 1993, pp. 1753-1767.
[63] Riazi, M. R. and Mansoori, G. A., "Simple Equation of State
Accurately Predicts Hydrocarbon Densities," Oil and Gas
Journal, July 12, 1993, pp. 108-111.
[64] Lee, B. I. and Kesler, M. G., "A Generalized Thermodynamic
Correlation Based on Three- Parameter Corresponding States,"
American Institute of Chemical Engineers Journal, Vol. 21, No. 5,
1975, pp. 510-527.
[65] Riazi, M. R. and Daubert, T. E., "Application of Corresponding
States Principles for Prediction of Self-Diffusion Coefficients in
Liquids," American Institute of Chemical Engineers Journal,
Vol. 26, No. 3, 1980, pp. 386-391.
[66] Riazi, M. R. and Mansoori, G. A., "Use of the Velocity of Sound
in Predicting the PVT Relations," Fluid Phase Equilibria, Vol. 90,
1993, pp. 251-264.
[67] Shabani, M. R., Riazi, M. R., and Shabau, H. I., "Use of Velocity
of Sound in Predicting Thermodynamic Properties from Cubic
Equations of State," Canadian Journal of Chemical Engineering,
Vol. 76, 1998, pp. 281-289.
[68] Riazi, M. R. and Roomi, Y. A., "Use of Velocity of Sound in
Estimating Thermodynamic Properties of Petroleum
Fractions,"Preprints of Division of Petroleum Chemistry
Symposia, American Chemical Society (ACS), August 2000,
Vol. 45, No. 4, pp. 661-664.
[69] Riazi, M. R. and Faghri, A., "Prediction of Thermal
Conductivity of Gases at High Pressures," American Institute of
Chemical Engineers Journal, Vol.31, No.l, 1985, pp. 164-167.
[70] Riazi, M. R. and Faghri, A., "Thermal Conductivity of Liquid
and Vapor Hydrocarbon Systems: Pentanes and Heavier at Low
Pressures," Industrial and Engineering Chemistry, Process
Design and Development, Vol. 24, No. 2, 1985, pp. 398-401.
[71] Ahmed, T., Hydrocarbon Phase Behavior, Gulf Publishing,
Houston, TX, 1989.
[72] Danesh, A., PVT and Phase Behavior of Petroleum Reservoir
Fluids, Elsevier, Amsterdam, 1998.
[73] Tsonopoulos, C., Heidman, J. L., and Hwang, S.-C.,
Thermodynamic and Transport Properties of Coal Liquids, An
Exxon Monograph, Wiley, New York, 1986.
[74] National Institute of Standards and Technology (NIST),
Boulder, CO, 2003 (http://webbook.nist.gov/chemistry/).
[75] Ahmed, T., Reservoir Engineering Handbook, Gulf Publishing,
Houston, TX, 2000.
[76] Edmister, W. C. and Lee, B. I., Applied Hydrocarbon
Thermodynamics, 2nd ed., Gulf Publishing, Houston, TX, 1985.
[77] HYSYS, "Reference Volume 1., Version 1.1", HYSYS Reference
Manual for Computer Software, HYSYS Conceptual Design,
Hyprotech Ltd., Calgary, Alberta, Canada, 1996.

[78] EPCON, "API Tech Database Software," EPCON International,
Houston, TX, 2000 (www.epcon.com).
[79] Stange, E. and Johannesen, S. O., "HYPO*S, A Program for
Heavy End Characterization," An Internal Program from the
Norsk Hydro Oil and Gas Group (Norway), Paper presented at
the 65th Annual Convection, Gas Processors Association, San
Antonio, TX, March 10, 1986.
[80] Aspen Plus, "Introductory Manual Software Version," Aspen
Technology, Inc., Cambridge, MA, December 1986.
[81] PRO/II, "Keyword Manual," Simulation Sciences Inc.,
Fullerton, CA, October 1992.
[82] Riazi, M. R., "Estimation of Physical Properties and
Composition of Hydrocarbon Mixures," Analytical Advartces
for Hydrocarbon Research, edited by Hsu, C. Samuel, Kluwer
Academic/Plenum Publishers, New York, NY, 2003, pp. 1-26.
[83] Riazi, M. R. and A1-Enezi, G., "A Mathematical Model for the
Rate of Oil Spill Disappearance from Seawater for Kuwaiti
Crude and Its Products," Chemical Engineering Journal, Vol. 73,
1999, pp. 161-172.
1. I N T R O D U C T I O N 29
[84] Riazi, M. R. and Edalat, M., "Prediction of the Rate of Oil
Removal from Seawater by Evaporation and Dissolution,"
Journal of Petroleum Science and Engineering, Vol. 16, 1996,
pp. 291-300.
[85] Fingas, M. E, "A Literature Review of the Physics
and Predictive Modeling of Oil Spill Evaporation,"
Journal of Hazardous Materials, Vol. 42, 1995, pp. 157-
175.
[86] Winn, E W., "Physical Properties by Nomogram," Petroleum
Refiners, Vol. 36, No. 21, 1957, pp. 157.
[87] Coplen, T. B., "Atomic Weights of the Elements 1999," Pure and
Applied Chemistry, Vol. 73, No. 4, 2001, pp. 667-683.
[88] Maddox, R. N., Gas Conditioning and Processing, Volume 4,
Campbell Petroleum Series, John M. Campbell, Norman, OK,
1985.
[89] Riazi, M. R., "Estimation of Rates and Enhancement Factors in
Gas Absorption with Zero-Order Reaction and Gas Phase Mass
Transfer Resistances," Chemical Engineering Science, Vol. 41,
No. 11, 1986, pp. 2925-2929.
 MNL50-EB/Jan. 2005

Characterization and
Properties of Pure
Hydrocarbons

NOMENCLATURE Greek Letters

API API gravity defined in Eq. (2.4)
A, B Parameters in a potential energy relation
A, B Parameters in a two-parameter cubic
equation of state
a, b, ... i Correlation constants in various
equations
CH Carbon-to-hydrogen weight ratio
d20 Liquid density at 20~ and 1 atm, g/cm 3
dc Critical density defined by Eq. (2.9), g/cm 3
F Intermolecular force
I Refractive index parameter defined in
Eq. (2.36)
Kw Watson (UOP) K factor defined by
Eq. (2.13)
In Natural logarithm (base e)
lOgl0 Logarithm to the base 10
M Molecular weight, g/rnol (kg/krnol) Superscript
n Sodium D line refractive index of
liquid at 20~ and latm, dimensionless
NA Avogadro's number
Nc Carbon number (number of carbon
Subscripts
atoms in a hydrocarbon molecule)
Pc Critical pressure, bar
pvap Vapor (saturation) pressure, bar
R Universal gas constant, 8.314 J/mol.K
/~ Refractivity intercept defined in Eq. (2.14)
Rm Molar refraction defined in Eq. (2.34),
crna/g
R 2 R squared, defined in Eq. (2.136)
r Distance between molecules
SG Specific gravity of liquid substance at
15.5~ (60~ defined by Eq. (2.2),
Acronyms
dimensionless
SGg Specific gravity of gas substance at 15.5~
(60~ defined by Eq. (2.6), dimensionless
Tb Boiling point, K API-TDB American Petroleum Institute--Technical Data
Tc Critical temperature, K
TF Flash point, K
TM Melting (freezing point) point, K
V Molar volume, cm3/gmol
V Saybolt universal viscosity, SUS
Vc Critical volume (molar), cm3/mol
(or critical specific volume, cm3/g)
VGC Viscosity gravity constant defined by
Eq. (2.15)
Zc Critical compressibility factor defined by
Eq. (2.8), dimensionless
P Potential energy defined in Eq. (2.19)
Polarizability defined by Eq. (2.33), cm3/mol
e Energy parameter in a potential energy relation
# Absolute (dynamic) viscosity, cp [mPa.s]. Also
used for dipole moment
u Kinematic viscosity defined by Eq. (2.12), cSt
[mm2/s]
0 A property of hydrocarbon such as M, To, Pc, Vc, I,
d, Tb . . . .
p Density at a given temperature and pressure,
g]cm 3
a Surface tension, dyn/cm (= raN/m)
a Size parameter in potential energy relation
w Acentric factor defined by Eq. (2.10),
dimensionless
~ Properties of n-alkanes from Twu correlations
A Aromatic
N Naphthenic
P Paraffinic
T Value of a property at temperature T
o A reference state for T and P
ec Value of a property at M -~ e~
20 Value of a property at 20~
38(100) Value of kinematic viscosity at 38~ (100~
99(210) Value of kinematic viscosity at 99~ (210~
%AAD Average absolute deviation percentage defined by
Eq. (2.135)
Book
%D Absolute deviation percentage defined by
Eq. (2.134)
EOS Equation of state
IUPAC International Union of Pure and Applied
Chemistry
%MAD Maximum absolute deviation percentage
NIST National Institute of Standards and Technology
RK Redlich-Kwong
vdW van der Waals
R 2 R squared, Defined in Eq. (2.136)

30

Copyright 9 2005 by ASTM International www.astm.org

 2. CHARACTERIZATION AND PROPERTIES OF PURE HYDROCARBONS
As DISCUSSEDIN CHAPTER1, the characterization of petroleum
fractions and crude oils depends on the characterization and
properties of pure hydrocarbons. Calculation of the prop-
erties of a mixture depends on the properties of its con-
stituents. In this chapter, first basic parameters and properties
of pure compounds are defined. These properties are either
temperature-independent or values of some basic properties
at a fixed temperature. These parameters are the basis for
calculation of various physical properties discussed in this
book. Reported values of these parameters for more than
100 selected pure compounds are given in Section 2.2. These
values will be used extensively in the following chapters, es-
pecially in Chapters 3 to determine the quality and properties
of petroleum fractions. In Section 2.3, the characterization of
hydrocarbons is introduced, followed by the development of
a generalized correlation for property estimation that is a
unique feature of this chapter. Various correlations and meth-
ods for the estimation of these basic parameters for pure hy-
drocarbons and narrow boiling range petroleum fractions are
presented in different sections. Finally, necessary discussion
and recommendations for the selection of appropriate pre-
dictive methods for various properties are presented.
2.1 DEFINITION O F BASIC PROPERTIES
In this section, all properties of pure hydrocarbons presented
in Section 2.2 are defined. Some specific characteristics of
petroleum products, such as cetane index and pour point, are
defined in Chapter 3. Definitions of general physical proper-
ties such as thermal and transport properties are discussed in
corresponding chapters where their estimation methods are
presented.
2.1.1 Molecular Weight
The units and definition of molecular weight or molar mass,
M, was discussed in Section 1.7.8. The molecular weight of a
pure compound is determined from its chemical formula and
the atomic weights of its elements. The atomic weights of the
elements found in a petroleum fluid are C = 12.011, H = 1.008,
S = 32.065, O = 16.0, and N = 14.01, as given by the IUPAC
standard [ 1]. As an example, the molecular weight of methane
(CH4) is calculated as 12.011 + 4 x 1.008 = 16.043 kg/kmol or
16.043 g/tool (0.01604 kg/mol) or 16.043 lb/lbmol. Molecular
weight is one of the characterization parameters for hydro-
carbons.
2.1.2 Boiling Point
The boiling point of a pure compound at a given pressure
is the temperature at which vapor and liquid exist together
at equilibrium. If the pressure is 1 atm, the boiling point is
called the normal boiling point. However, usually the term
boiling point, Tb, is used instead of normal boiling point and
for other pressures the term saturation temperature is used.
In some cases, especially for heavy hydrocarbons in which
thermal cracking may occur at high temperatures, boiling
points at pressures other than atmospheric is specified. Boil-
ing points of heavy hydrocarbons are usually measured at 1,
10, or 50 m m Hg. The conversion of boiling point from low
31
pressure to normal boiling point requires a vapor pressure re-
lation and methods for its calculation for petroleum fractions
are discussed in Chapter 3. The boiling point, when available,
is one of the most important characterization parameters for
hydrocarbons and is frequently used in property estimation
methods.
2.1.3 Density, Specific Gravity, and API Gravity
Density is defined as mass per unit volume of a fluid. Density
is a state function and for a pure compound depends on both
temperature and pressure and is shown by p. Liquid densities
decrease as temperature increases but the effect of pressure
on liquid densities at moderate pressures is usually negligible.
At low and moderate pressures (less than a few bars), satu-
rated liquid density is nearly the same as actual density at the
same temperature. Methods of the estimation of densities of
fluids at various conditions are discussed in Chapters 5 and 7.
However, liquid density at the reference conditions of 20~
(293 K) and 1 atm is shown b y d and it is used as a characteri-
zation parameter in this chapter as well as Chapter 3. Parame-
ter d is also called absolute density to distinguish from relative
density. Other parameters that represent density are specific
volume (l/d), molar volume (M/d), and molar density (d/M).
Generally, absolute density is used in this book as the charac-
teristic parameter to classify properties of hydrocarbons.
Liquid density for hydrocarbons is usually reported in
terms of specific gravity (SG) or relative density defined as
density of liquid at temperature T
(2.1) SG =
density of water at temperature T
Since the standard conditions adopted by the petroleum in-
dustry are 60 ~F (15.5 ~C) and 1 atm, specific gravities of liquid
hydrocarbons are normally reported at these conditions. At a
reference temperature of 60~ (15.5~ the density of liquid
water is 0.999 g/cm 3 (999 kg/m 3) or 8.337 lb/gal(U.S.). There-
fore, for a hydrocarbon or a petroleum fraction, the specific
gravity is defined as
(2.2) SG (60~176 -- density of liquid at 60~ in g/cm 3
0.999 g/cm 3
Water density at 60~ is 0.999 or almost 1 g/cm3; therefore,
values of specific gravities are nearly the same as the density
of liquid at 15.5~ (289 K) in g/cm 3. The Society of Petroleum
Engineers usually uses y for the specific gravity and in some
references it is designated by S. However, in this book SG de-
notes the specific gravity. Since most of hydrocarbons found
in reservoir fluids have densities less than that of water, spe-
cific gravities of hydrocarbons are generally less than 1. Spe-
cific gravity defined by Eq. (2.2) is slightly different from the
specific gravity defined in the SI system as the ratio of the den-
sity of hydrocarbon at 15~ to that of water at 4~ designated
b y d 15. Note that density of water at 4~ is exactly I g/cm 3 and
therefore d15 is equal to the density of hydrocarbon at 15~
in g/cm 3. The relation between these two specific gravities is
approximately given as follows:
(2.3) SG = 1.001 d4is
In this book specific gravity refers to SG at 60~176 (15.5~
In the early years of the petroleum industry, the American
32 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
Petroleum Institute (API) defined the API gravity (degrees
API) to quantify the quality of petroleum products and crude
oils. The API gravity is defined as [2]
141.5
(2.4) API gravity - SG (at 60~ - 131.5
Degrees API was derived from the degrees Baum6 in which
it is defined in terms of specific gravity similar to Eq. (2.4)
except numerical values of 140 and 130 were used instead
of 141.5 and 131.5, respectively. Liquid hydrocarbons with
lower specific gravities have higher API gravity. Aromatic hy-
drocarbons have higher specific gravity (lower API gravity)
than do paraffinic hydrocarbons. For example, benzene has
SG of 0.8832 (API of 28.72) while n-hexane with the same car-
bon number has SG of 0.6651 (API gravity of 81.25). A liquid
with SG of i has API gravity of 10. Once Eq. (2.4) is reversed it
can be used to calculate specific gravity from the API gravity.
141.5
(2.5) SG =
API gravity + 131.5
The definition of specific gravity for gases is somewhat dif-
ferent. It is defined as relative density of gas to density of air at
standard conditions. In addition, density of gases is a strong
function of pressure. Since at the standard conditions (15.5 ~C
and 1 arm) the density of gases are estimated from the ideal
gas law (see Chapter 5), the specific gravity of a gas is pro-
portional to the ratio of molecular weight of gas (Mg) to the
molecular weight of air (28.97).
(2.6) SGg-- Mg
28.97
Therefore, to obtain the specific gravity of a gas, only its
molecular weight is needed. For a mixture, Mg can be deter-
mined from the gas composition, as discussed in Chapter 3.
2.1.4 R e f r a c t i v e Index
Refractive index or refractivity for a substance is defined as
the ratio of velocity of light in a vacuum to the velocity of
light in the substance (fluid) and is a dimensionless quantity
shown by n:
velocity of light in the vacuum
(2.7) n=
velocity of light in the substance
In other words, when a light beam passes from one substance
(air) to another (a liquid), it is bent or refracted because of
the difference in speed between the two substances. In fact,
refractive index indicates the degree of this refraction. Refrac-
tive index is a state function and depends on the temperature
and pressure of a fluid. Since the velocity of light in a fluid is
less than the velocity of light in a vacuum, its value for a fluid
is greater than unity. Liquids have higher values of refractive
index than that of gases. For gases the values of refractive
index are very close to unity.
All frequencies of electromagnetic radiation (light) travel
at the same speed in vacuum (2.998 x l0 s m/s); however, in a
substance the velocity of light depends on the nature of the
substance (molecular structure) as well as the frequency of
the light. For this reason, standard values of refractive index
must be measured at a standard frequency. Usually the refrac-
tive index of hydrocarbons is measured by the sodium D line
at 20~ and 1 atm. The instrument to measure the refractive
index is called a refractometer and is discussed in Chapter 3.
In some references the values of refractive index are reported
at 25~ however, in this book the refractive index at 20~ and
1 arm is used as a characterization parameter for hydrocar-
bons and petroleum fractions. As is shown in this chapter and
Chapter 3, refractive index is a very useful characterization
parameter for pure hydrocarbons and petroleum fractions,
especially in relation with molecular type composition. Val-
ues of n vary from about 1.3 for propane to 1.6 for some
aromatics. Aromatic hydrocarbons have generally higher
n values than paraffinic compounds as shown in Table 2.1.
2.1.5 Critical Constants (T~, Pc, V~, Zc)
The critical point is a point on the pressure-volume-
temperature diagram where the saturated liquid and satu-
rated vapor are identical and indistinguishable. The temper-
ature, pressure, and volume of a pure substance at the critical
point are called critical temperature (To), critical pressure (Pc),
and critical volume (Vc), respectively. In other words, the crit-
ical temperature and pressure for a pure compound are the
highest temperature and pressure at which the vapor and liq-
uid phase can coexist at equilibrium. In fact, for a pure com-
pound at temperatures above the critical temperature, it is
impossible to liquefy a vapor no matter how high the pres-
sure is. A fluid whose temperature and pressure are above
the critical point is called supercritical fluid. For pure com-
pounds, critical temperature and pressure are also called true
critical temperature and true critical pressure. However, as
will be discussed in Chapter 3, pseudocritical properties are
defined for mixtures and petroleum fractions, which are dif-
ferent from true critical properties. Pseudocritical properties
are important in process calculations for the estimation of
thermophysical properties of mixtures.
The critical compressibility factor, Z~, is defined from To, Pc,
and Vc according to the general definition of compressibility
factor.
PcVc
(2.8) Zc =
RTc
where R is the universal gas constant. According to Eq. (2.8),
Zc is dimensionless and Vc must be in terms of molar vol-
ume (i.e., cm3/mol) to be consistent with R values given in
Section 1.7.24. Critical temperature, pressure, and volume
( Tc, Pc, Vc) are called the critical constants or critical properties.
Critical constants are important characteristics of pure com-
pounds and mixtures and are used in corresponding states
correlations and equations of state (EOS) to calculate PVT and
many other thermodynamic, physical, and transport proper-
ties. Further discussion on the critical point of a substance
is given in Chapter 5. As was discussed in Section 1.3, the re-
sults of EOS calculations very much depend on the values of
critical properties used. Critical volume may be expressed in
terms of specific critical volume (i.e., ma/kg), molar critical
volume (i.e., ma/kinol), or critical density dc (i.e., kg/m 3) or
critical molar density (i.e., kmol/m3). Critical density is re-
lated to the critical molar volume as
M
(2.9) d~ = - -
vo
 2. C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P U R E H Y D R O C A R B O N S
Experimental values of critical properties have been re-
ported for a large number of pure substances. However, for
hydrocarbon compounds, because of thermal cracking that
occurs at higher temperatures, critical properties have been
measured up to C18 [2]. Recently some data on critical proper-
ties of n-alkanes from C19 to C36 have been reported [3]. How-
ever, such data have not yet been universally confirmed and
they are not included in major data sources. Reported data
on critical properties of such heavy compounds are generally
predicted values and vary from one source to another. For ex-
ample, the API-TDB [2] reports values of 768 K and 11.6 bar
for the critical temperature and pressure of n-eicosane, while
these values are reported as 767 K and 11.1 bar by Poling
et al. [4]. Generally, as boiling point increases (toward heav-
ier compounds), critical temperature increases while critical
pressure decreases. As shown in Section 2.2, aromatics have
higher Tc and Pc relative to those of paraffinic compounds
with the same carbon atoms.
2.1.6 Acentric Factor
Acentric factor is a parameter that was originally defined by
Pitzer to improve accuracy of corresponding state correla-
tions for heavier and more complex compounds [5, 6]. Acen-
tric factor is a defined parameter and not a measurable quan-
tity. It is a dimensionless parameter represented by w and is
defined as
(2.10) w = - log10 (PIvap) -- 1.0
where
pvap = reduced vapor pressure, pvap/Pc, dimensionless
pvap = vapor pressure at T = 0.7 Tc (reduced temperature
of 0.7), bar
Pc -- critical pressure, bar
T = absolute temperature, K
Tc = critical temperature, K
Acentric factor is defined in a way that for simple fluids such
as argon and xenon it is zero and its value increases as the
size and shape of molecule changes. For methane w -- 0.001
and for decane it is 0.489. Values reported for acentric fac-
tor of pure compounds are calculated based on Eq. (2.10),
which depends on the values of vapor pressure. For this rea-
son values reported for the acentric factor of a compound may
slightly vary from one source to another depending on the re-
lation used to estimate the vapor pressure. In addition, since
calculation of the acentric factor requires values of critical
temperature and pressure, reported values for w also depend
on the values of Tc and Pc used.
2.1.7 Vapor Pressure
In a closed container, the vapor pressure of a pure compound
is the force exerted per unit area of walls by the vaporized
portion of the liquid. Vapor pressure, pvap, can also be de-
fined as a pressure at which vapor and liquid phases of a
pure substance are in equilibrium with each other. The vapor
pressure is also called saturation pressure, psat, and the cor-
responding temperature is called saturation temperature. In
an open air under atmospheric pressure, a liquid at any tem-
perature below its boiling point has its own vapor pressure
that is less than 1 atm. When vapor pressure reaches 1 atm,
33
the saturation temperature becomes the normal boiling point.
Vapor pressure increases with temperature and the highest
value of vapor pressure for a substance is its critical pressure
(Pc) in which the corresponding temperature is the critical
temperature (To). When a liquid is open to the atmosphere at
a temperature T in which the vapor pressure of liquid is pvap,
vol% of the compound vapors in the air is
(2.11) vol% = 100 • \ Pa ]
where Pa is the atmospheric pressure. Derivation of Eq. (2.11)
is based on the fact that vapor pressure is equivalent to partial
pressure (mole fraction • total pressure) and in gases under
low-pressure conditions, mole fraction and volume fraction
are the same. At sea level, where P~ = 1 atm, calculation of
vol% of hydrocarbon vapor in the air from Eq. (2.11) is simply
100 pvap, if pwp is in atm.
Vapor pressure is a very important thermodynamic prop-
erty of any substance and it is a measure of the volatility of
a fluid. Compounds with a higher tendency to vaporize have
higher vapor pressures. More volatile compounds are those
that have lower boiling points and are called light compounds.
For example, propane (Ca) has boiling point less than that of
n-butane (nCa) and as a result it is more volatile. At a fixed
temperature, vapor pressure of propane is higher than that
of butane. In this case, propane is called the light compound
(more volatile) and butane the heavy compound. Generally,
more volatile compounds have higher critical pressure and
lower critical temperature, and lower density and lower boil-
ing point than those of less volatile (heavier) compounds, al-
though this is not true for the case of some isomeric com-
pounds. Vapor pressure is a useful parameter in calculations
related to hydrocarbon losses and flammability of hydrocar-
bon vapor in the air (through Eq. 2.11). More volatile com-
pounds are more ignitable than heavier compounds. For ex-
ample, n-butane is added to gasoline to improve its ignition
characteristics. Low-vapor-pressure compounds reduce evap-
oration losses and chance of vapor lock. Therefore, for a fuel
there should be a compromise between low and high vapor
pressure. However, as will be seen in Chapter 6, one of the
major applications of vapor pressure is in calculation of equi-
librium ratios (Ki values) for phase equilibrium calculations.
Methods of calculation of vapor pressure are given in detail in
Chapter 7. For pure hydrocarbons, values of vapor pressure at
the reference temperature of 100~ (38~ are provided by the
API [2] and are given in Section 2.2. For petroleum fractions,
as will be discussed in Chapter 3, method of Reid is used to
measure vapor pressure at 100~ Reid vapor pressure (RVP)
is measured by the ASTM test method D 323 and it is approx-
imately equivalent to vapor pressure at 100~ (38~ RVP is
a major characteristic of gasoline fuel and its prediction is
discussed in Chapter 3.
2.1.8 Kinematic Viscosity
Kinematic viscosity is defined as the ratio of absolute (dy-
namic) viscosity/z to absolute density p at the same temper-
ature in the following form:
(2.12) v = ~-
P
34 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
As discussed in Section 1.7.18, kinematic viscosity is ex-
pressed in cSt, SUS, and SFS units. Values of kinematic vis-
cosity for pure liquid hydrocarbons are usually measured and
reported at two reference temperatures of 38~ (100~ and
99~ (210~ in cSt. However, other reference temperatures
of 40~ (104~ 50~ (122~ and 60~ (140~ are also used
to report kinematic viscosities of petroleum fractions. Liq-
uid viscosity decreases with an increase in temperature (see
Section 2.7). Kinematic viscosity, as it is shown in Chapter 3,
is a useful characterization parameter, especially for heavy
fractions in which the boiling point may not be available.
2.1.9 Freezing and Melting Points
Petroleum and most petroleum products are in the form of a
liquid or gas at ambient temperatures. However, for oils con-
taining heavy compounds such as waxes or asphaltinic oils,
problems may arise from solidification, which cause the oil
to lose its fluidity characteristics. For this reason knowledge
of the freezing point is important and it is one of the ma-
jor specifications of jet fuels and kerosenes. For a pure com-
pound the freezing point is the temperature at which liquid
solidifies at 1 atm pressure. Similarly the melting point, TM,
is the temperature that a solid substance liquefies at 1 atm.
A pure substance has the same freezing and melting points;
however, for petroleum mixtures, there are ranges of melting
and freezing points versus percent of the mixture melted or
frozen. For a mixture, the initial melting point is close to the
melting point of the lightest compound in the mixture, while
the initial freezing point is close to the freezing point (or melt-
ing point) of the heaviest compound in the mixture. Since the
melting point increases with molecular weight, for petroleum
mixtures the initial freezing point is greater than the initial
melting point. For petroleum mixtures an equivalent term of
pour point instead of initial melting point is defined, which
will be discussed in Chapter 3. Melting point is an important
characteristic parameter for petroleum and paraffinic waxes.
2.1.10 Flash Point
Flash point, TF, for a hydrocarbon or a fuel is the minimum
temperature at which vapor pressure of the hydrocarbon is
sufficient to produce the vapor needed for spontaneous igni-
tion of the hydrocarbon with the air with the presence of an
external source, i.e., spark or flame. From this definition, it is
clear that hydrocarbons with higher vapor pressures (lighter
compounds) have lower flash points. Generally flash point
increases with an increase in boiling point. Flash point is
an important parameter for safety considerations, especially
during storage and transportation of volatile petroleum prod-
ucts (i.e., LPG, light naphtha, gasoline) in a high-temperature
environment. The surrounding temperature around a storage
tank should always be less than the flash point of the fuel
to avoid possibility of ignition. Flash point is used as an in-
dication of the fire and explosion potential of a petroleum
product. Estimation of the flash point of petroleum fractions
is discussed in Chapter 3, and data for flash points of some
pure hydrocarbons are given in Table 2.2. These data were ob-
tained using the closed cup apparatus as described in ASTM
D 93 (ISO 2719) test method. There is another method of
measuring flash point known as open cup for those oils with
flash point greater than 80~ (ASTM D 92 or ISO 2592 test
methods). Flash point should not be mistaken with fire point,
which is defined as the minimum temperature at which the
hydrocarbon will continue to burn for at least 5 s after being
ignited by a flame.
2.1.11 Autoignition Temperature
This is the minimum temperature at which hydrocarbon va-
por when mixed with air can spontaneously ignite without
the presence of any external source. Values of autoignition
temperature are generally higher than flash point, as given
in Table 2.2 for some pure hydrocarbons. Values of autoigni-
tion temperature for oils obtained from mineral sources are
in the range of 150-320~ (300-500~ for gasoline it is about
350~ (660~ and for alcohol is about 500~ (930~ [7].
With an increase in pressure the autoignition temperature
decreases. This is particularly important from a safety point
of view when hydrocarbons are compressed.
2.1.12 F l a m m a b i l i t y R a n g e
To have a combustion, three elements are required: fuel (hy-
drocarbon vapor), oxygen (i.e., air), and a spark to initiate the
combustion. One important parameter to have a good com-
bustion is the ratio of air to hydrocarbon fuel. The combustion
does not occur if there is too much air (little fuel) or too lit-
tle air (too much fuel). This suggests that combustion occurs
when hydrocarbon concentration in the air is within a certain
range. This range is called flammability range and is usually
expressed in terms of lower and upper volume percent in the
mixture of hydrocarbon vapor and air. The actual volume per-
cent of hydrocarbon vapor in the air may be calculated from
Eq. (2.1 i) using vapor pressure of the hydrocarbon. If the cal-
culated vol% of hydrocarbon in the air is within the flamma-
bility range then the mixture is flammable by a spark or flame.
2.1.13 Octane Number
Octane number is a parameter defined to characterize an-
tiknock characteristic of a fuel (gasoline) for spark ignition
engines. Octane number is a measure of fuel's ability to re-
sist auto-ignition during compression and prior to ignition.
Higher octane number fuels have better engine performance.
The octane number of a fuel is measured based on two ref-
erence hydrocarbons of n-heptane with an assigned octane
number of zero and isooctane (2,2,4-trimethylpentane) with
assigned octane number of 100. A mixture of 70 vol% isooc-
tane and 30 vol% n-heptane has an octane number of 70.
There are two methods of measuring octane number of a fuel
in the laboratory. The methods are known as motor octane
number (MON) and research octane number (RON). The MON
is indicative of high-speed performance (900 rpm) and is mea-
sured under heavy road conditions (ASTM D 357). The RON
is indicative of normal road performance under low engine
speed (600 rpm) city driving conditions (ASTM D 908). The
third type of octane number is defined as posted octane num-
ber (PON), which is the arithmetic average of the MON and
RON [PON = (MON + RON)/2]. Generally isoparaffins have
higher octane number than do normal paraffins. Naphthenes
have relatively higher octane number than do corresponding
 2. C H A R A C T E R I Z A T I O N
paraffins and aromatics have very high octane numbers. The
octane number of a fuel can be improved by adding tetra-
ethyl-lead (TEL) or methyl-tertiary-butyl-ether (MTBE). Use
of lead (Pb) to improve octane number of fuels is limited in
many industrial countries. In these countries MTBE is used
for octane number improvement. However, there are prob-
lems of groundwater contamination with MTBE. MTBE has
MON and RON of 99 and 115, respectively [8]. Lead gener-
ally improves octane number of fuels better than MTBE. The
addition of 0.15 g Pb/L to a fuel of RON around 92 can im-
prove its octane number by 2-3 points. With 0.6 g Pb/L one
may improve the octane number by 10 points [8]. However,
as mentioned above, because of environmental hazards use
of lead is restricted in many North American and West Euro-
pean countries. Values of the octane number measured with-
out any additives are called clear octane number. For pure
hydrocarbons values of clear MON and RON are given in Sec-
tion 2.2. Estimation of the octane number of fuels is discussed
in Chapter 3.
2.1.14 Aniline Point
The aniline point for a hydrocarbon or a petroleum fraction is
defined as the minimum temperature at which equal volumes
of liquid hydrocarbon and aniline are miscible. Aniline is an
aromatic compound with a structure of a benzene molecule
where one atom of hydrogen is replaced by the -NH2 group
(C6Hs-NH2). The aniline point is important in characteriza-
tion of petroleum fractions and analysis of molecular type.
As discussed in Chapter 3, the aniline point is also used as a
characterization parameter for the ignition quality of diesel
fuels. It is measured by the ASTM D 611 test method. Within
a hydrocarbon group, aniline point increases with molecu-
lar weight or carbon number, but for the same carbon num-
ber it increases from aromatics to paraffinic hydrocarbons.
Aromatics have very low aniline points in comparison with
paraffins, since aniline itself is an aromatic compound and it
has better miscibility with aromatic hydrocarbons. Generally,
oils with higher aniline points have lower aromatic content.
Values of the aniline point for pure hydrocarbons are given
in Table 2.2, and its prediction for petroleum fractions is dis-
cussed in Chapter 3.
2.1.15 Watson K
Since the early years of the petroleum industry it was desired
to define a characterization parameter based on measurable
parameters to classify petroleum and identify hydrocarbon
molecular types. The Watson characterization factor denoted
by Kw is one of the oldest characterization factors originally
defined by Watson et al. of the Universal Oil Products (UOP)
in mid 1930s [9]. For this reason the parameter is sometimes
called UOP characterization factor and is defined as
(1.8Tb) 1/3
(2.13) Kw =
SG
where
Tb = normal boiling point K
SG = specific gravity at 15.5~
In the original definition of Kw, boiling point is in degrees
Rankine and for this reason the conversion factor of 1.8
is used to have Tb in the SI unit. For petroleum fractions
AND PROPERTIES OF PURE HYDROCARBONS 35
Tb is the mean average boiling point (also see Chapter 3). The
purpose of definition of this factor was to classify the type
of hydrocarbons in petroleum mixtures. The naphthenic hy-
drocarbons have Kw values between paraffinic and aromatic
compounds. In general, aromatics have low Kw values while
paraffins have high values. However, as will be discussed in
Chapter 3 there is an overlap between values of Kw from dif-
ferent hydrocarbon groups. The Watson K was developed in
1930s by using data for the crude and products available in
that time. Now the base petroleum stocks in general vary sig-
nificantly from those of 1930s [10, 11]. However, because it
combines two characterization parameters of boiling point
and specific gravity it has been used extensively in the devel-
opment of many physical properties for hydrocarbons and
petroleum fractions [2, 11, 12].
2.1.16 Refractivity Intercept
Kurtz and Ward [13] showed that a plot of refractive index
against density for any homologous hydrocarbon group is
linear. For example, plot of refractive index of n-paraffins ver-
sus density (d20) in the carbon number range of C5-C45 is a
straight line represented by equation n = 1.0335 + 0.516d20,
with R E value of 0.9998 (R2 = I, for an exact linear relation).
Other hydrocarbon groups show similar performance with
an exact linear relation between n and d. However, the inter-
cept for various groups varies and based on this observation
they defined a characterization parameter called refractivity
intercept, R~, in the following form:
d
(2.14) Ri = n - -
2
where n and d are refractive index and density of liquid hy-
drocarbon at the reference state of 20~ and I atm in which
density must be in g/cm 3. Ri is high for aromatics and low for
naphthenic compounds, while paraffins have intermediate/~
values.
2.1.17 Viscosity Gravity Constant
Another parameter defined in the early years of petroleum
characterization is the viscosity gravity c o n s t a n t (VGC). This
parameter is defined based on an empirical relation developed
between Saybolt viscosity (SUS) and specific gravity through
a constant. VGC is defined at two reference temperatures of
38~ (100~ and 99~ (210~ as [14]
10SG - 1.0752 log10(V38 - 38)
(2.15) VGC =
10 - logl0(V3s - 38)
SG - 0.24 - 0.022 log10(V99 - 35.5)
(2.16) VGC =
0.755
where
V38 = viscosity at 38~ (100~ in SUS (Saybolt Universal
Seconds)
V99 = Saylbolt viscosity (SUS) at 99~ (210~
Conversion factors between cSt and SUS are given in Sec-
tion 1.7.18. Equations (2.15) and (2.16) do not give identical
values for a given compound but calculated values are close to
each other, except for very low viscosity oils. Equation (2.16)
is recommended only when viscosity at 38~ (100~ is not
36 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
available. VGC varies for paraffinic hydrocarbons from 0.74 to
0.75, for naphthenic from 0.89 to 0.94, and for aromatics from
0.95 to 1.13 [15]. In Chapter 3, VGC along with other param-
eters has been used to estimate the composition of petroleum
fractions. Values of VGC for some hydrocarbons are given in
Table 2.3. The main limitation in use of VGC is that it cannot
be defined for compounds or fractions with viscosities less
than 38 SUS (--~3.6 cSt) at 38~ A graphical method to esti-
mate VGC of petroleum fractions is presented in Chapter 3.
ASTM D 2501 suggests calculation of VGC using specific
gravity and viscosity in mm2/s (cSt) at 40~ v40 in the follow-
ing form:
SG - 0.0664 - 0.1154 logl0(v40 - 5.5)
(2.17) VGC--
0.94 - 0.109 logl0(v40 - 5.5)
Values of VGC calculated from Eq. (2.17) are usually very
close to values obtained from Eq. (2.15). If viscosity at 40~
is available, use of Eq. (2.17) is recommended for calculation
of VGC. Another relation to calculate VGC in metric units was
proposed by Kurtz et al. [16] in terms of kinematic viscosity
and density at 20~ which is also reported in other sources
[17].
(2.18) VGC= d - 0.1384 log10(v20-20) +0.0579
0.1526[7.14 - logl0(V20 - 20)]
in which d is density at 20~ and 1 atm in g/cm 3 and v20 is the
kinematic viscosity at 20~ in cSt. In this method viscosity
of oil at 20~ must be greater than 20 cSt. However, when
there is a choice Eq. (2.15) should be used for the procedures
described in Chapter 3.
Example 2.1--API RP-42 [ 18] reports viscosity of some heavy
hydrocarbons, 1,1-Di-(alphadecalyl)hendecane (C31H56) is a
naphthenic compound with molecular weight of 428.8 and
specific gravity of 0.9451. The kinematic viscosity at 38~
(100~ is 20.25 cSt. Calculate the viscosity gravity constant
for this compound.
Solution--Using Eq. (1.17), the viscosity is converted from
cSt to SUS: V3s = 99.5 SUS. Substituting values of V3s and
SG = 0.9451 into Eq. (2.15) gives VGC = 0.917. The VGC may
be calculated from Eq. (2.17) with direct substitution of vis-
cosity in the cSt unit. Assuming there is a slight change in
viscosity from 38 to 40~ the same value of viscosity at 38~
is used for v40. Thus v40 ~ 20.25 cSt (mm2/s) and Eq. (2.17)
gives VGC = 0.915. The small difference between calculated
values of VGC because in Eq. (2.17) viscosity at 40~ must
be used, which is less than the viscosity at 38~ Calculated
VGC is within the range of 0.89-0.94 and thus the hydrocar-
bon must be a naphthenic compound (also see Fig. 3.22 in
Chapter 3). #
2.1.18 Carbon-to-Hydrogen Weight Ratio
Carbon-to-hydrogen weight ratio, CH weight ratio, is defined
as the ratio of total weight of carbon atoms to the total weight
of hydrogen in a compound or a mixture and is used to
characterize a hydrocarbon compound. As was discussed in
Section 1.1.1, hydrocarbons from different groups have dif-
ferent formulas. For example, alkanes (paraffins) have the
general formula of CnHan+2, alkylcyclopentanes or alkylcy-
clohexanes (naphthenes) have formula of C, H2,, and alkyl-
benzenes (aromatics) have formula of CnH2n-6 (n ~ 6). This
shows that at the same carbon number, the atomic ratio of
number of carbon (C) atoms to number of hydrogen (H)
atoms increases from paraffins to naphthenes and aromat-
ics. For example, n-hexane (C6H14), cyclohexane (C6H12), and
benzene (C6H6)from three different hydrocarbon groups all
have six carbon atoms, but have different CH atomic ratios
of 6/14, 6/12, and 6/6, respectively. If CH atomic ratio is mul-
tiplied by the ratio of atomic weights of carbon (12.011) to
hydrogen (1.008), then CH weight ratio is obtained. For ex-
ample, for n-hexane, the CH weight ratio is calculated as
(6/14)x(12.011/1.008) = 5.107, This number for benzene is
11.92. Therefore, CH weight ratio is a parameter that is ca-
pable of characterizing the hydrocarbon type. In addition,
within the same hydrocarbon group, the CH value changes
from low to high carbon number. For example, methane has
CH value of 2.98, while pentane has CH value of 4.96. For
extremely large molecules (M ~ o0), the CH value of all hy-
drocarbons regardless of their molecular type approaches the
limiting value of 5.96. This parameter is used in Section 2.3
to estimate hydrocarbon properties, and in Chapter 3 it is
used to estimate the composition of petroleum fractions. In
some references HC atomic ratio is used as the character-
izing parameter. According to the definition, the CH weight
ratio and HC atomic ratio are inversely proportional. The
limiting value of HC atomic ratio for all hydrocarbon types
is 2.
Another use of CH weight ratio is to determine the quality of
a fossil-type fuel. Quality and the value of a fuel is determined
from its heat of combustion and heating value. Heating value
of a fuel is the amount of heat generated by complete com-
bustion of 1 unit mass of the fuel. For example, n-hexane has
the heating value of 44734 kJ/kg (19232 Btu/lb) and benzene
has the heating value of 40142 kJ/kg (17258 Btu/lb). Calcula-
tion of heating values are discussed in Chapter 7. From this
analysis it is clear that as CH value increases the heating value
decreases. Hydrogen (H2), which has a CH value of zero, has a
heating value more than that of methane (CH4) and methane
has a heating value more than that of any other hydrocarbon.
Heavy aromatic hydrocarbons that have high CH values have
lower heating values. In general, by moving toward lower CH
value fuel, not only do we have better heating value but also
better and cleaner combustion of the fuel. It is for this reason
that the use of natural gas is preferable to any other type of
fuel, and hydrogen is an example of a perfect fuel with zero
CH weight ratio (CH = 0), while black carbon is an example
of the worst fuel with a CH value of infinity. Values of CH for
pure hydrocarbons are given in Section 2.2 and its estimation
methods are given in Section 2.6.3.
2.2 DATA O N BASIC P R O P E R T I E S
OF S E L E C T E D P U R E H Y D R O C A R B O N S
2.2.1 Sources o f Data
There are several sources that provide data for physical prop-
erties of pure compounds. Some of these sources are listed
below.
 2. CHARACTERIZATION AND PROPERTIES OF PURE HYDROCARBONS
1. API: Technical Data Book~Petroleum Refining [2]. The first
chapter of API-TDB compiles basic properties of more than
400 pure hydrocarbons and some nonhydrocarbons that
are important in petroleum refining. For some compounds
where experimental data are not available, predicted values
from the methods recommended by the API are given.
2. DIPPR: Design Institute for Physical and Property Data
[ 19]. The project initially supported by the AIChE began in
early 1980s and gives various physical properties for both
hydrocarbon and nonhydrocarbon compounds important
in the industry. A computerized version of this data bank is
provided by EPCON [20].
3. TRC Thermodynamic Tables--Hydrocarbons [21]. The
Thermodynamic Research Center (formerly at the Texas
A&M University) currently at the National Institute of Stan-
dards and Technology (NIST) at Boulder, CO, (http://www.
trc.nist.gov/) in conjunction with the API Research Project
44 [22] has regularly published physical and basic proper-
ties of large number of pure hydrocarbons.
4. API Research Project 44 [22]. This project sponsored by the
API was conducted at Texas A&M University and provides
physical properties of selected hydrocarbons.
5. API Research Project 42 [18]. This data compilation com-
pleted in the 1960s provides experimental data on den-
sity, refractive index, viscosity, and vapor pressure for
more than 300 hydrocarbons with carbon number greater
than Cu.
6. Dortmund Data Bank (DDB) [23]. This project on physical
properties has been conducted at the University of Olden-
burg in Germany. DDB contains experimental data from
open literature on various thermodynamic properties of
pure compounds and some defined mixtures. Data have
been programmed in a computer software convenient for
extracting data. Majority of data are on thermodynamic
properties, such as vapor-liquid equilibrium (VLE), activ-
ity coefficients, and excess properties. However, data on
viscosity, density, vapor pressure, thermal conductivity, and
surface tension have also been complied as mentioned in
their Web site. Unfortunately they have not compiled char-
acteristic data on hydrocarbons and petroleum fractions
important in the petroleum industry. Also the data on trans-
port properties are mainly for pure compounds at atmo-
spheric pressures.
7. The fourth and fifth editions of The Properties of Gases
and Liquids [4] also provide various properties for more
than 400 pure compounds (hydrocarbons and nonhydro-
carbons). However, data in this book have been mainly
taken from the TRC Tables [21 ].
8. There are also some free online sources that one may use to
obtain some physical property data. The best example is the
one provided by NIST (http://webbook.nist.gov). Various
universities and researchers have also developed special on-
line sources for free access to some data on physical proper-
ties. For example the Center for Research in Computational
Thermochemistry (CRCT) of Ecole Ploytechnique Mon-
treal provides online calculational software at http://www.
crct.polymfl.caJfact/index.php/. The Center for Applied
Thermodynamic Studies (CATS) at the University of Idaho
also provides softwares for property calculations at its web-
site (http://www.webpages.uidaho.edu/~cats/). G. A. Man-
soori in his personal Web site also provides some online
37
sources for physical property data (http://tigger.uic.edu/
~mansoori/Thermodynamic.Data.and.Property-html/).
In m a n y occasions different sources provide different values
for a particular property depending on the original source
of data. Calculated properties such as critical constants and
acentric factor for compounds heavier than C18 should be
taken with care as in different sources different methods have
been used to predict these parameters.
2.2.2 Properties o f Selected Pure Compounds
The basic properties of pure hydrocarbons from different
groups that will be used in the predictive methods presented
in the following chapters are tabulated in Tables 2.1 and 2.2.
The basic properties of M, TM, Tb, SG, d20, n20, To, Pc, Vc, Zc,
and ~0are presented in Table 2.1. Secondary properties of kine-
matic viscosity, API gravity, Kw, vapor pressure, aniline point,
flash and autoignition points, flammability range, and oc-
tane number are given in Table 2.2. Compounds selected are
mainly hydrocarbons from paraffins, naphthenes, and aro-
matics that constitute crude oil and its products. However,
some olefinic and nonhydrocarbons found with petroleum
fluids are also included. Most of the compounds are from ho-
mologous hydrocarbon groups that are used as model com-
pounds for characterization of petroleum fractions discussed
in Chapter 3. The properties tabulated are the basic prop-
erties needed in characterization techniques and thermody-
namic correlations for physical properties of petroleum frac-
tions. Although there are separate chapters for estimation of
density, viscosity, or vapor pressure, these properties at some
reference temperatures are provided because of their use in
the characterization methods given in Chapter 3 and 4. Other
physical properties such as heat capacity or transport prop-
erties are given in corresponding chapters where the predic-
tive methods are discussed. Data for more than 100 selected
compounds are presented in this section and are limited to
C22 mainly due to the lack of sufficient experimental data for
heavier compounds. Data presented in Tables 2.1 and 2.2 are
taken from the API-TDB [2, 22]. Standard methods of mea-
surement of these properties are presented in Chapter 3.
Example 2.2--Assume large amount of toluene is poured on
the ground in an open environment at which the temperature
is 38~ (100~ Determine if the area surrounding the liquid
surface is within the flammability range.
Solution--From Table 2.2, the flammability range is 1.2-7.1
vol% of toluene vapor. From this table, the vapor pressure of
toluene at 38~ is 0.071 bar (0.07 atm). Substituting this vapor
pressure value in Eq. (2.11) gives the value of vol% = 100 x
0.07/1.0 = 7% of toluene in the air mixture. This number is
within the flammability range (1.2 < 7 < 7.1) and therefore
the surrounding air is combustible. r
2.2.3 Additional Data on Properties
o f H e a v y Hydrocarbons
Some data on density, refractive index, and viscosity of some
heavy hydrocarbons are given in Table 2.3. These data are
taken from API RP 42 [18]. Values of P~ and VGC in the table
 TABLE 2.1--Basic physical properties of some selected compounds.
TM, Tb, SG, d2o, TC, Pc, VC,
Compound Formula Nc M ~ ~ at 60~ g/ore3 n20 ~ bar cm3/mol Zc
Paraffins
Methane CH4 1 16.0 -182.5 -161.5 0.2999 ... -82.59 45.99 98.65 0.2864 0.0115
Ethane C2H6 2 30.1 -182.8 -88.6 0.3554 0.33"86 ... 32.17 48.72 145.48 0.2792 0.0995
Propane C3H8 3 44,1 -187.7 -42.0 0.5063 0.4989 ... 96.68 42.48 200.14 0.2765 0.1523
n-Butane C4H10 4 58.1 -138.3 -0.5 0.5849 0.5791 1.3326 151.97 37.96 255.09 0.2740 0.2002
n-Pentane C5H12 5 72.2 -129.7 36.1 0.6317 0.6260 1.3575 196.55 33.70 313.05 0.2702 0.2515
n-Hexane C6H14 6 86.2 -95.3 68.7 0.6651 0.6605 1.3749 234.45 30.25 371.22 0.2661 0.3013
n-Heptane C7H16 7 100.2 -90.6 98.4 0.6902 0.6857 1.3876 267.05 27.40 427.88 0.2610 0.3495
n-Octane C8H18 8 114.2 -56.8 125.7 0.7073 0.7031 1.3974 295.55 24.90 486.35 0.2561 0.3996
n-Nonane C9H20 9 128.3 -53.5 150.8 0.7220 0.7180 1.4054 321.45 22.90 543.67 0,2519 0.4435
n-Decane C10H22 10 142,3 -29.6 174.2 0.7342 0.7302 1.4119 344.55 21.10 599.58 0.2463 0.4923
n-Undecane Cl1H24 11 156.3 -25.6 195.9 0.7439 0.7400 1.4151 365.85 19.50 658.69 0,2418 0.5303
n-Dodecane C12H26 12 170,3 -9.6 216.3 0.7524 0.7485 1.4195 384.85 18.20 715.67 0.2381 0.5764
n-Tridecane C13H28 13 184.4 -5.4 235.5 0.7611 0.7571 1.4235 401.85 16.80 774.60 0.2319 0.6174
n-Tetradecane C14H30 14 198.4 5.9 253.6 0.7665 0.7627 1.4269 419.85 15.70 829.82 0.2261 0.6430
n-Pentadecane C15H32 15 212.4 9.9 270.7 0.7717 0.7680 1.4298 434.85 14.80 888.49 0.2234 0.6863
t~ n-Hexadecane C16H34 16 226.4 18.2 286.9 0.7730 0.7729 1.4325 449.85 14.00 944.33 0.2199 0.7174
n-Heptadecane C17H36 17 240.5 22.0 302.2 0.7752 0.7765 1.4348 462.85 13.40 999.83 0.2189 0.7697
n-Octadecane C18H38 18 254.5 28.2 316.7 0.7841 0.7805 1.4369 473.85 12.70 1059.74 0,2167 0.8114
n-Nonadecane C19H40 19 268.5 31.9 329.9 0.7880 0.7844 1.4388 484.85 12.10 1119.82 0.2150 0.8522
n-Eicosane C20H42 20 282.6 36.4 343.8 0.7890 0.7871 1.4405 494.85 11.60 1169.50 0.2125 0.9069
n-Heneicosane C21H44 21 296.6 40.2 356.5 0.7954 0.7906 1.4440 504.85 11.10 1229.41 0.2110 0.9420
n-Docosane C22H46 22 310.6 44.0 368.6 0.7981 0.7929 1.4454 513.85 10,60 1289.49 0.2089 0.9722
isobutane (2-Methylpropane) C4H10 4 58.1 -159.6 -11.7 0.5644 0.5584 134.99 36.48 262.71 0.2824 0.1808
isopentane (2-Methylbutane) CsHI2 5 72.2 -159.9 27.8 0,6265 0.6213 1.3"5"37 187.28 33.81 305.84 0.2701 0.2275
2-Methylpentane C6H14 6 86.2 -95.3 60.3 0.6577 0.6529 1.3715 224.35 30.05 371.22 0.2661 0.2774
2-Methylhexane C7H16 7 100.2 -118.3 90.1 0.6822 0.6778 1,3849 257.22 27.34 421.00 0.2610 0.3277
2-Methylheptane C8H18 8 114.2 -109.0 117.7 0.7029 0.6987 1.3949 286.49 24.84 487.78 0.2604 0.3772
2,2,4-Trimethylpentane (isooctane) C8H18 8 114.2 -107.4 99.2 0.6988 0.6945 1.3915 270.81 25.68 467.81 0.2656 0.3022
2-Methyloctane C9H20 9 128.3 -80.4 143.3 0.7176 0.7134 1.4031 313.60 22.90 541.27 0.2541 0.4212
2-Methylnonane C10H22 10 142.3 -74.7 167.0 0.7307 0.7266 1.4100 336.85 21.20 582.70 0.2436 0.4723
Olefins
Ethene (Ethylene) C2H4 2 28.1 -169.2 -103.7 0.1388 ... 9.19 50.40 131.00 0.2813 0.0865
Propene (Propylene) C3H6 3 42.1 -185.3 -47.7 0.5192 0.5'1"11 ... 92.42 46.65 188.36 0,2891 0.1398
1-Butene C4H8 4 56.1 -185.4 -6.3 0.6001 0.5938 146.80 40.43 239.24 0.2770 0.1905
1-Pentene CsHt0 5 70,1 -165.2 30.0 0.6456 0,6402 1.3715 191.63 35.13 295.10 0.2683 0.2312
1-Hexene C6H12 6 84.2 -139.8 63.5 0.6790 0.6740 1.3879 230.88 31.40 354.13 0.2654 0.2804
1-Heptene C7H14 7 98.2 -118.9 93.6 0,7015 0.6971 1.3998 264.14 28.30 413.15 0.2617 0.3310
1-Octene C8H16 8 112.2 -101.7 121.3 0.7181 0.7140 1.4087 293.50 25.68 460.26 0.2509 0.3764
1-Nonene C9H18 9 126.2 -81.4 146.9 0.7330 0.7290 ... 320.10 23.30 528.04 0.2494 0.4171
1-Decene C10H20 10 140.3 -66.3 170.6 0.7450 0.7410 ... 343.25 22.18 584.08 0.2528 0.4800
Acetylene C2H2 2 26.0 -80.8 -83.8 ... 0.4001 ... 35.17 61.39 112.97 0.2706 0.1873
1,3-Butadiene C4H6 4 54.1 -108.9 -4.4 ... 0.6219 ... 152.02 42.77 220.49 0.2668
 Naphthenes
Cyclopentane CsH10 5 70.1 -93.8 49.3 0.7502 0.7456 1.4065 238.61 45.02 257.89 0.2729 0.1959
Methylcyclopentane C6H12 6 84.2 -142.4 71.8 0.7540 0.7491 1.4097 259.64 37.85 318.92 0.2725 0.2302
Ethylcyclopentane C7H14 7 98.2 -138.4 103.5 0.7712 0.7667 1.4198 296.37 33.98 375.14 0.2692 0.2716
Propylcyclopentane C8H16 8 112.2 -117.3 131.0 0.7811 0.7768 1.4263 322.85 30.20 428.03 0.2609 0.3266
n-Butylcyclopemane C9H18 9 126.2 -108.0 156.6 0.7893 0.7851 1.4316 347.85 27.20 483.11 0.2545 0.3719
n-Pentylcyclopentane C10H20 10 140.3 -83.0 180.5 0.7954 ... 1.4352 370.65 24.50 536.79 0.2457 0,4184
n-Hexylcyclopentane CllH22 11 154.3 -73.0 202.9 0.8006 ... 1.4386 390.95 22.20 592.40 0.2382 0.4646
n-Heptylcyclopentane C12H24 12 168.3 -53.0 223.9 0.8051 ... 1.4416 409.45 20.10 647.30 0.2293 0,5100
n-Octylcyclopentane C13H26 13 182.4 -44.0 243.5 0.8088 ... 1.4446 426.35 18.30 702.38 0.2210 0.5525
n-Nonylcyclopentane C14H28 14 196.4 -29.0 262.0 0.8121 ... 1.4467 441.75 16.70 757.64 0.2129 0,5956
n-Decylcyclopentane C15H30 15 210.4 -22.1 279.4 0.8149 ... 1.4486 455.95 15.30 811.76 0.2049 0,6314
n -Undecylcyclopentane C16H32 16 224.4 -10.0 295.8 0.8175 ... 1.4503 469.05 14.00 867,27 0.1968 0.6741
n-Dodecylcyclopentane C17H34 17 238.5 -5.0 311.2 0.8197 ... 1.4518 481.25 12.80 921.48 0.1881 0.7163
n-Tridecylcyclopentane C18H36 18 252.5 5.0 325.9 0.8217 ... 1.4531 492.45 11.80 977.26 0.1812 0.7582
n -Tetradecylcyclopentane CIgHa8 19 266.5 9.0 340.0 0.8235 ... 1.4543 502.85 10.90 1031,55 0.1743 0,7949
n-Pemadecylcyclopentane C20H40 20 280.5 17.0 353.0 0.8252 ... 1.4554 512.55 10.00 1087.59 0.1665 0.8395
n -Hexadecylcyclopentane C21H42 21 294,6 21.0 366.0 0.8267 ... 1.4564 521.55 9.20 1141,97 0.1590 0.8755
n-Heptadecylcyclopentane C22H44 22 308.6 27.0 377.0 0.8280 1.4573 538.01 11.91 1198.28 0.2115 0.9060
Cyclohexane C6H12 6 84.2 6.5 80.7 0.7823 0.8021 1.4262 280.43 40.73 307.89 0.2725 0.2096
Methylcyclohexane C7H14 7 98.2 -126.6 100.9 0,7748 0.7702 1.4231 299.04 34.71 367,79 0.2684 0,2350
Ethylcyclohexane C8H16 8 112.2 -111.3 131.8 0.7926 0.7884 1.4330 336.00 30.40 430.13 0.2582 0.2455
Propylcyclohexane C9H18 9 126.2 -94.9 156.8 0.7981 0.7940 1.4371 366.00 28.07 476,81 0.2519 0.2595
n-Butylcyclohexane C10H20 10 140.3 -74.7 181.0 0.8033 0.7993 1.4408 393.85 25.70 534.16 0.2476 0.2743
t~ Aromatics
Benzene C6H 6 6 78.1 5.5 80.1 0.8832 0.8780 1.5011 289.01 48.98 258.94 0.2714 0,2100
Methylbenzene (Toluene) C7H 8 7 92.1 -95.0 110.6 0.8741 0.8685 1.4969 318.65 41.06 315.80 0.2635 0.2621
Ethylbenzene C8H10 8 106.2 -95.0 136.2 0.8737 0.8678 1.4959 344.05 36.06 373.81 0.2627 0.3026
Propylbenzene C9H12 9 120.2 -99.6 159.2 0.8683 0.8630 1.4920 365.23 32.00 439.71 0.2651 0.3447
n-Butylbenzene C10H14 10 134.2 -87.9 183.3 0.8660 0.8610 1.4898 387.40 28.87 496.89 0.2612 0.3938
n-Pentylbenzene CllH16 11 148.2 -75.0 205.5 0.8624 0.8583 1.4878 406.75 26.04 549.74 0.2532 0,4378
n-Hexylbenzene ClaHl8 12 162.3 --61.2 226.1 0.8622 0.8581 1.4864 424.85 23.80 592.64 0.2431 0.4790
n-Heptylbenzene C13H20 13 176.3 -48.0 246.1 0.8617 0.8576 1.4854 440.85 21.80 648,27 0.2381 0.5272
n-Octylbenzene C14H22 14 190.3 -36.0 264.4 0.8602 0.8562 1.4845 455.85 20.20 703.41 0.2344 0.5670
n-Nonytbenzene C15H24 15 204.4 -24.2 282.1 0.8596 0.8557 1.4838 467.85 18.95 752,70 0.2315 0.6331
n-Decylbenzene C16H26 16 218.4 -14.4 297.9 0.8590 0.8551 1.4832 479.85 17.70 812.55 0.2297 0.6797
n-Undecylbenzene C17H28 17 232.4 -5.2 313.3 0.8587 0.8548 1.4828 490.85 16.72 867.64 0.2284 0.7333
n-Dodecylbenzene C18H30 18 246.4 2.8 327.6 0.8595 0.8556 1.4824 506.85 15.60 923.08 0.2221 0.7333
n -Tridecylbenzene C19H32 19 260.5 10.0 341.3 0.8584 0.8545 1.4821 516.85 14.80 977.25 0.2202 0.7799
n-Tetradecylbenzene C20H34 20 274.5 16.0 354.0 0.8587 0.8553 1.4818 526.85 14.00 1029,88 0.2168 0.8130
n-Pentadecylbenzene C21H36 21 288.5 22.0 366.0 0.8587 0.8540 1.4815 535.85 13.30 1089.71 0.2155 0.8567
n-Hexadecylbenzene C22H38 22 302.5 27.0 378.0 0.8586 0.8541 1.4813 544.85 12.70 1140,80 0.2130 0.8996
o-xylene (o-dimethylbenzene) C8H10 8 106.2 -25.2 144.4 0.8849 0,8799 .,, 357.18 37.34 369.17 0.2630 0.3104
m-xylene CsH10 8 106.2 -47.9 139.1 0.8691 0.8643 1.4972 343.90 35.36 375,80 0.2590 0.3259
(Continued)
 TABLE 2.1--(Continued)
TM, Tb, SG, d20, Tc, Pc, Vct
Compound Formula Nc M ~ ~ at 60~ g/cm 3 n20 ~ bar cm3/mol Zc
p-xylene C8H10 8 106.2 13.3 138.4 0.8654 0.8608 1.4958 343.08 35.11 379.11 0.2598 0.3215
isopropylbenzene C9H12 9 120.2 -96.0 152.4 0.8682 0.8632 1.4915 357.95 32.09 426.95 0.2611 0.3258
1,2,3-Trimethylbenezene C9H12 9 120.2 -25.4 176.1 0.8985 0.8942 1.5130 391.38 34.54 414.20 0.2590 0.3664
isobutylbenzene C10H14 10 134.2 -51.5 172.8 0.8577 0.8532 1.4865 377.00 30.40 455.83 0.2564 0.3797
o-Cymene C10H14 10 134.2 -71.5 178.2 0.8812 0.8765 1.5005 388.85 29.30 489.35 0.2605 0.3372
Cyclohexylbenzene C12H16 12 160.3 7.0 240.1 0.8617 0.9430 ... 470.85 28.80 529.25 0.2464 0.3783
Naphthalene C10H8 10 128.2 80.3 218.0 1.0281 1.0238 ... 475.20 40.51 412.89 0.2688 0.3022
1-Methylnaphthalene CllH10 11 142.2 -30.5 244.7 1.0242 1.1020 ... 498.89 36.60 465.17 0.2652 0.3478
Styrene C8H8 8 104.2 -30.6 145.2 0.9097 0.9049 ... 362.85 38.40 351.76 0.2555 0.2971
Diphenyl methane C13H12 13 168.2 25.2 264.3 1.0101 1.0244 ... 494.85 29.20 547.20 0.2502 0.4615
Nonhydrocarbons
O Water H20 ... 18.0 0.0 100.0 1.0000 0.9963 ... 373.98 220.55 55.91 0.2292 0.3449
C a r b o n dioxide CO2 1 44.0 -56.6 -78.5 0.8172 0.7745 ... 31.06 73.83 93.96 0.2743 0.2236
H y d r o g e n Sulfide H2S ... 34.1 -85.5 -60.4 0.8012 0.7901 ... 100.38 89.63 98.51 0.2843 0.0942
Nitrogen N2 ... 28.0 -210.0 -195.8 0.8094 . . . . . . -146.95 34.00 89.21 0.2891 0.0377
Oxygen 02 ... 32.0 -218.8 -183.0 1.1420 . . . . . . -118.57 50.43 73.32 0.2877 0.0222
Ammonia NH3 ... 17.0 -77.7 -33.4 0.6165 0.6093 ... 132.50 112.80 72.51 0.2425 0.2526
Carbon monoxide CO 1 28.0 -205.0 -191.5 . . . . . . . . . -140.23 34.99 94.43 0.2990 0.0482
Hydrogen H2 ... 2.0 -259.2 -252.8 . . . . . . . . . -239.96 13.13 64.28 0.3059 -0.2160
Air N2+O2 ... 29.0 -214.0 -194.5 -140.70 37.74 91.45 0.3134 0.0074
Methanol CHaOH 1 32.04 -97.7 64.7 0.7993 0.7"944 1.3265" 239.49 80.97 118.00 0.2240 0.5640
Ethanol C2HsOH 2 46.07 -114.1 78.29 0.7950 0.7904 1.3594" 240.77 61.48 167.00 0.2400 0.6452
Isobutanol C4H9OH 4 74.12 -108.0 107.66 0.8063 0.8015 1.3938" 274.63 43.02 272.97 0.2580 0.5848
Nc: carbon number; d20: liquid density at 20~ nz0: refractive index at 20~ (at 25~ TM: freezing point; Tb: normal boiling point; Tc: critical temperature;
Pc: critical pressure; Vc: critical volume; Zc: critical compressibility factor; to: acentric factor. Data source: API-TDB [2].
*SG of light gases are at low temperatures where the gas is in liquid form.
 TABLE 2.2---Secondary properties o f some selected compounds.
p,sat Flammability
API at 100~ Viscosity, cSt TF AI, AP, range, vol.% RON
Compound Formula Arc gravity Kw bar v380oo) V99(210) CH ~ ~ ~ Min Max clear
Paraffins
Methane CH4 1 340,3 19.53 344.737 . . . . . . 2.98 . . . . . . 536.9 5.00 1.00
Ethane C2H6 2 266.6 19.49 55.1579 ... 3.97 . . . . . . 471.9 2.90 13.00
Propane C3H8 3 148.0 14.74 12.9621 0.1775 . . 4.47
. . . ... . 449.9 2.10 9.50 1.8
n-Butane C4H10 4 110.4 13.49 3,5608 0.2533 0.1686 4.77 -.. 83.2 287.9 1.80 8.40 93.8
n-Pentane C5H12 5 92.5 13.02 1.0745 0.3394 0.2643 4.96 --40.0 70.8 242.9 1.40 8.30 61.7
n-Hexane C6H14 6 81.2 12.79 0.3435 0.4152 ... 5.11 --21.7 68.7 224.9 1.20 7.70 24.8
n-Heptane C7H16 7 73.5 12.67 0.1112 0.5046 0.3515 5.21 --4.1 ... 203.9 1.00 7.00 0.0
n-Octane C8H18 8 68.6 12.66 0.0370 0.6364 0.3997 5.30 12.9 ... 205.9 0.96
n-Nonane C9H20 9 64.5 12.66 0.0125 0.8078 0.4694 5.36 30.9 ... 204.9 0.87 2.90
n-Decane CloH22 10 61.2 12.67 0.0042 1.0154 0.5537 5.42 45.9 ... 200.9 0.78 2.60
n-Undecane CllH24 11 58.7 12.71 0.0014 1.2588 0.6397 5.46 65.0 ... 201.9
n-Dodecane C12H26 12 56.6 12.74 0.0005 1.5452 0.7469 5.50 73.9 ... 202.9
n-Tridecane C13H28 13 54.4 12.76 0.0002 1.8634 0.8624 5.53 78.9 ... 201.9
4~ n-Tetradecane C14H30 14 53.1 12.82 0.0001 2.2294 0.9885 5.56 100.0 ... 199.9
n-Pentadecane C15H32 15 51.9 12.87 0.0000 2.6415 1.1328 5.59 113.9 ... 201.9
n-Hexadecane C16H34 16 51.6 12.97 0.0000 3.1229 1.2859 5.61 135.1 ... 201.9
n-Heptadecane C17H36 17 51.0 13.05 0.0000 3.6045 1.4413 5.63 147.8 ... 201.9
n-Octadecane C18H38 18 49.0 13.01 0.0000 4.1620 1.5815 5.64 165.1 ... 201.9
n-Nonadecane C 19H40 19 48.1 13.04 0.0000 4.6090 1.7940 5.66 167.9 ... 201.9
n-Eicosane C20H42 20 47.8 13.12 0.0000 5.3165 1.9889 5.67 166.9 ... 201.9
n-Heneicosane C21H44 21 46.4 13.11 ... ... 2.1703 5.69 176.9 ... 201.9
n-Docosane C22H46 22 45.8 13.15 ... 2.4099 5.70 184.9 201.9
isobutane (2-Methylpropane) C4H10 4 119.2 13.78 5.0199 0.2"773 0.1873 4.77 - - H 107".7 460.1 1.80 8.40 8;1
i s o p e n t a n e (2-Methylbutane) C5H12 5 94.4 13.01 1.4110 0.3066 ... 4.96 -57.2 70.8 420.1 1.40 8.30 92.3
2-Methylpentane C6H14 6 83.6 12.82 0.4666 0.3862 .,. 5.11 -35.2 73,8 ... 1.20 7.70 73.4
2-Methylhexane C7H16 7 75.9 12.72 0.1562 0.4730 ... 5.21 -23.2 74.1 ... 1.00 7.00 42.4
2-Methylheptane C8H18 8 69.8 12.65 0.0528 0.5908 0.3635 5.30 4.1 73.9 ,.. 0.98 ... 20.6
2,2,4-Trimethylpentane (isooctane) C8H18 8 71.0 12.52 0.1181 0.6077 0.3738 5.30 -12.2 79.5 ... 1.00 ... 100.0
2-Methyloctane C9H20 9 65.7 12.66 0.0177 0.7382 0.4329 5.36 22.9 . . . . . .
2-Methylnonane C10H22 10 62.1 12.66 0.0058 0.9636 0.5401 5.42 40.9 . . . . . .
(Continued)
 TABLE 2.2--(Continued)
p,sat Flammability
API at 100~ Viscosity, cSt TF AI, AP, range, vol. % RON
Compound Formula Arc gravity Kw bar v38(100) v99(210) CH ~ ~ ~ Min Max clear
Olefins
Ethene (Ethylene) C2H4 2 888.0 48.49 . . . . . . 5.96 2.30 32.30
Propene (Propylene) C3H 6 3 141.0 14.26 15.7812 0.1"801 ... 5.96 -108.2 455.1 2.00 11.00 100.2
1-Butene C4H8 4 104.3 13.05 4.2953 . . . . . . 5.96 383.9 1.60 9.30 97.4
1-Pentene CsH10 5 87.7 12.66 1.3203 ... 5.96 -i8".2 19.1 272.9 1.50 8.70 90.9
1-Hexene C6H12 6 76.9 12.46 0.4143 0.3415 ... 5.96 -31.2 22.8 253.1 1.00 7.50 76.4
1-Heptene C7H14 7 70.2 12.41 0.1353 0.4317 0.3043 5.96 0.1 27.3 263.1 0.80 6.90 54.5
1-Octene C8H16 8 65.5 12.42 0.0453 0.5657 0.3590 5.96 20.9 32.6 230.1 0.80 6.80 28.7
1-Nonene C9H18 9 61.5 12.43 0.0152 0.7004 0.4294 5.96 26.9 38.1 236,9 0.60 6.00
1-Decene CmH20 10 58.4 12.45 0.0051 0.8848 0.5027 5.96 47.1 44,2 235.1 0.55 5.70
Acetylene C2H2 2 . . . . . . . . . . . . 11.92 -18.2 305.1 2.50 80.00
1,3 Butadiene C4H6 4 . . . . . . 4.0906 0.2030 0.2248 7.94 428.9
Naphthenes
Cyclopentane CsH10 5 57.1 11.12 0.6839 0.4973 .,. 5.96 -39.2 16.8 361.1 1.40 9.40 100.1
Methylcyclopentane C6H12 6 56.2 11.31 0.3106 0.5646 ... 5,96 -27.2 33.1 328.9 1.20 8.35 91.4
Ethylcyclopentane C7H14 7 52.0 11.39 0.0970 0.6199 .., 5.96 -4.1 36.8 260.1 1.10 6.70 67.2
Propylcyclopentane C8H~6 8 49.7 11.51 0.0325 0.7257 0.4613 5.96 15.9 44.5 269.1 0.95 6.40 31.2
n-Butylcyclopentane C9H18 9 47.8 11.63 0.0108 0.9118 0.5148 5.96 31.9 48.8 250.1 0.80 5.90 97.0
n-Pentylcyclopentane CloH20 10 46.4 11.75 ... 1,1280 0.6200 5.96 0.74 5.47
4~ n-Hexylcyclopentane CllH22 11 45.2 11.86 ... 1.4150 0.7300 5.96 0.68 5.20
t~
n-Heptylcyclopentane C12H24 12 44.3 11.97 .., 1.7480 0.8500 5.96 0.62 5.06
n-Octylcyclopentane C13H26 13 43.5 12.07 ... 2.1300 0.9800 5.96 0.57 5.01
n-Nonylcyclopentane C14H28 14 42.7 12.16 ... 2.5700 1.1200 5.96 0.53 5.07
n-Decylcyclopentane C15H30 I5 42.1 12.25 ... 3.0500 1.2700 5.96 0.50 5.24
n-Undecylcyclopentane C16H32 16 41.6 12.33 ... 3.6300 1.4400 5.96 0.47 5.53
n-Dodecylcyclopentane C17H34 17 41.1 12.41 ... 4.2500 1.6100 5.96 0.44 5.95
n-Tridecylcyclopentane C18H36 18 40.7 12.48 ... 4.9500 1.7800 5.96 0.41 6.53
n-Tetradecylcyclopentane C19H38 19 40.3 12.55 ... 5.7100 1.9800 5.96 0.39 7.33
n-Pentadecylcyclopentane C20H40 20 40.0 12.61 ... 6.5600 2.1900 5.96 0.37 8.38
n-Hexadecylcyclopentane C21H42 21 39.7 12.67 ... 7.4900 2.4000 5.96 0.35 9.79
n-Heptadecylcyclopentane C22H44 22 39.4 12.73 . . . . . . 5.96 0.34 11.67
Cyclohexane C6H12 6 49,4 11.00 0,2"274 0.9419 ... 5.96 -ik0 31.1 26;.1 1.30 8.00 83.o
Methylcyclohexane C7H14 7 51.1 11.31 0.1106 0.7640 0.4757 5.96 -5.9 41.1 285.1 1.15 7.20 74.8
Ethylcyclohexane C8H16 8 47.0 11.35 0.0333 0.8629 0.5122 5.96 22.1 43.8 261.9 0.90 6.60 45.6
Propylcyclohexane C9H18 9 45.8 11.50 0.0117 1.0010 0.5759 5.96 30.9 49.8 248.1 0.95 5.90 17.8
n-Butylcydohexane C10H20 10 44.6 11.64 0.0040 1.2539 0.6100 5.96 47.9 54.4 246.1 0.85 5.50
Aromatics
Benzene C6H6 6 28,7 9.74 0.2216 0.5927 0.3306 11.92 -11.2 -30.0 560.1 1.40 7.10
Methylbenzene (Toluene) C7H8 7 30.4 10.11 0.0710 0.5604 0.3433 10.43 4.9 -30.0 480.1 1.20 7.10 lOS'.8
Ethylbenzene C8H1o 8 30.5 10.34 0.0257 0.6540 0.3970 9.53 15.1 -30.0 430.1 1.00 6.70 100.8
Propylbenzene C9H12 9 31.5 10.59 0.0100 0.7977 0.4534 8.94 30.1 -30.0 456,1 0.88 6.00 101.5
n-Butylbenzene CloH14 10 31.9 10.82 0.0033 0.9483 0.5186 8.51 50.1 -30.0 410.1 0.80 5.80 100.4
 n-Pentylbenzene CllH16 11 32.6 11,03 0.0011 1.1824 0.6295 8.19 65.1 . . . . . . 0.80 5.50 ...
n-Hexylbenzene C12H18 12 32.6 11.19 0.0004 1.4419 0.7334 7.94 80.1 0.70 5.30
n-Heptylbenzene C13H20 13 32.7 11.35 0.0001 1.7546 0.8546 7.75 95.1 . . . . . . 0.70 5.10 ...
n-Octylbenzene C14H22 14 33.0 11.50 0.0000 2.0974 0.9763 7.58 107.1 . . . . . . 0.70 4.90 ...
n-Nonylbenzene C15H24 15 33.1 11.63 0.0000 2.5329 1.1089 7.45 98.9 . . . . . . 0.60 4.70 ...
n-Decylbenzene C16H26 16 33.2 11.75 0.0000 3.0119 1.2490 7.33 106.9 . . . . . . 0.60 4.60 ...
n-Undecylbenzene C17Hz8 17 33.3 11.86 0.0000 3.5672 1.3983 7.23 143.9 . . . . . . 0.60 4.40 ...
n-Dodecylbenzene C18H30 18 33.1 11.94 0.0000 4.1857 1.5509 7.15 140.9 . . . . . . 0.60 4.30 ...
n-Tridecylbenzene C19H32 19 33.3 12.05 0.0000 4.9566 1.7258 7.07 164.9 . . . . . . 0.50 4.20 ...
n-Tetradecylbenzene C20H34 20 33.3 12.13 0.0000 5.7107 1.8853 7.01 173.9 . . . . . . 0.50 4.10 ...
n-Pentadecylbenzene C21H36 21 33.3 12.20 0.0000 6.5716 2.0862 6.95 182.9 . . . . . . 0.50 4.00 ...
n-Hexadecylbenzene C22H38 22 33,3 12.28 0.0000 7.4138 2.2939 6.90 192.9 ... 0.50 3.90 ...
o-xylene ( o - d i m e t h y l b e n z e n e ) C8H10 8 28.4 10.27 0.0182 0.4238 9.53 16.9 463". 1 1.00 6.00
m-xylene CsHIo 8 31.3 10.42 0.0226 0.5936 0.3644 9.53 25.1 -3'0.0 465.1 1.10 7.00 104.0
p-xylene CsH10 8 32.0 10.46 0.0237 0.6167 0.3704 9.53 25.1 -30.0 528.1 1.10 7.00 103.4
isopropylbenzene C9H12 9 31.5 10.54 0.0129 0.7474 0.4229 8.94 43.9 -15.0 ... 0.88 6.50 102.1
1,2,3-Trimethylbenezene C9H12 9 26.0 10.37 0.0049 0.8317 0.3979 8.94 51.2 . . . . . . 0.88 5.20 101.4
isobutylbenzene CIoH14 10 33.5 10.84 0.0057 0.9895 0.5166 8.51 55.1 . . . . . . 0.80 6.00 101.6
o-Cymene CloH14 10 29.1 10.59 0.0045 0.9739 0.5979 8.51 53.1 . . . . . . 0,80 5.20 100.6
Cyclohexylbenzene C12H16 12 32.7 11.30 0.0002 2.0086 0.7936 8.94 98.9 . . . . . . 0.70 5.40 ...
Naphthalene C10H8 10 6.1 9.34 . . . . . . 0.7747 14.89 80.1 ... 526.1 0.88 5.90 ...
1-Methylnaphthalene CllH10 11 6.7 9.54 0.0002 2.1837 0.9204 13.11 82.1 ... 0.80 5.30
Styrene C8H8 8 24.0 10.00 0.0170 0.6637 0.3783 11.92 31.9 ... 490.1 1.10 6.10 10"3'.0
DiphenyI m e t h a n e C13H12 13 8.6 9.79 0.0001 2.1883 0.9834 12.91 130.1 . . . . . . 0.70 5.20 ...
Nonhydrocarbons
Water H20 . 10.0 8.76 0.0655 0.7063 0.0003 . . . . . . . . . . . . . . . . . . . . .
C a r b o n dioxide CO2 "i ... 8.63 . . . . . . . . . . . . . . . . . . . . .
Hydrogen Sulfide H2S ...... 9.06 27.2295 0.i447 . . . . . . . . . . . . 260.1 4.3"0 45.50 ...
Nitrogen N2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Oxygen 02 ...... 4.78 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ammonia NH3 ... 12.26 14.5632 0.1990 0.1373 . . . . . . . . . 650.9 16.00 25.00 ..,
Carbon monoxide CO "'i . . . . . . . . . . . . . . . - . . . . . . . . . 608.9 12.50 74.00 ...
Hydrogen H2 . . . . . . . . . . . . . . . . . . . . . . . . . . . 400.1 4.00 75.00 ...
Air N2+O2 . . . . . . . . . . . . . . . . . . . . . . . . .
Methanol CH3OH "1 45~52 10.60 0.3189 0.5'900 0.3269 2.98 lb185 463.85 ... 2~3 36.0 130
Ethanol C2H5OH 2 46.48 10.80 0.1603 1.0995 0.4579 3.97 12.85 422.85 ... 4.3 19.0 125
Isobutanol C4H9OH 4 44.0 10.94 0.0335 2.8700 0.7527 4.77 27.85 . . . . . . 1.7 10.9 ...
Nc: carbon number; psat : true vapor pressure at 100~ (311 K); v: kinematic viscosity; CH: carbon-to-hydrogen weight ratio; TM: freezing point; AI: autoignition temperature; AP: aniline point; RON: research
octane number. Data sources: API-TDB [2]; for autoignition temperature: DIPPR [20].
 TABLE 2.3--Additional data on refractive index and viscosity of heavy hydrocarbons.
Components Formula M SG d v V VGC
Paraffins
n-Tetracosane C24H50 338.6 0.8028 0.7980 2.79 35.5 1.4504 1.051 0.745
11-n-Butyldocosane C26H54 366.7 0.8081 0.8041 2.78 35.5 1.4461 1.044 0.749
5-n-Bytyldocosane C26H54 366.7 0.8099 0.8059 3.02 36.3 1.4466 1.044 0.750
9-Ethyl-9-n-Heptyloctadecane C27H56 380.7 0.8117 0.8076 3.48 37.8 1.4513 1.048 0.746
7-n-Hexyldocosane C28H58 394.7 0.8121 0.8080 3.32 38.2 1.4517 1.048 0.749
11-n-Decylhenecicosane C31H64 436.8 0.8158 0.8117 3.66 38.4 1.454 1.084 0.7481
n-Dotriacontane C32H66 450.8 0.8160 0.8119 5.01 42.7 1.455 1.049 0.730
11-n-Decyldocosane C32H66 450.8 0.8170 0.8129 3.91 39.2 1.4543 1.048 0.747
1 l-n-Decyltetracosane C34H70 478.9 0.8194 0.8153 4.45 40.9 1.4556 1.048 0.746
9-n-Octylhexacosane C38H78 478.9 0.8187 0.8156 4.70 14.7 1.456 1.049 0.743
13-n-Dodeecylhexacosane C38H78 507.0 0.8209 0.8168 4.90 42.4 1.4567 1.040 0.745
13-n-Undecylpentacosane C38H78 535.0 0.8230 0.8189 5.54 4.44 1.4577 1.048 0.744
Naphthenes
1,3 -Dicylohexylcyclohexane C18H32 248.4 0.9490 0.940 4.61 41.4 1.5065 a 1.036 0.917
1,2-Dicylohexylcyclohexane C18H32 248.4 0.949 0.9404 4.86 42.2 1.5062 1.036 0.917
2(ar)-n-Butyl-3 (ar)-n-hexyl-tetralin C20H32 272.5 0.9182 0.9086 2.92 35.9 1.5075 1.048 0.909
Perhydridibenzo-(a,i) fluorene C21H34 286.5 1.0081 0.9981 10.66 61.6 1.5277 1.0287 0.976
Di(alpha-decalyl)-methane C21H36 288.5 0.9807 0.9709 10.96 62.6 1.518 1.031 0.940
1,2-Di(alpha-decalyl)ethane C22H38 302.5 0.9748 0.9661 12.2 67.2 1.5176 1.034 0.930
1,1-Di(alpha-decalyl)ethane C22H38 302.5 0.9854 0.9765 16.5 83.9 1.5217 1.033 0.938
1,5-Dicyclopentyl-3(2-cyclopentylethyl)p e n t a n e C22H38 304.5 0.9040 0.895 3.96 39.5 1.4853 1.0378 0.866
1,7-Dicyclopentyl-4(2-phenylethy) heptane C25H40 340.6 0.9298 0.9205 3.92 39.2 1.507 1.046 0.899
1-Phenyl-3 (2-cyclohexylethyl)-6-cyclopentylhexane C25H40 340.6 0.9318 0.9230 4.80 42 1.5086 1.0465 0.893
6-n-Ocylperhydrobenz(de) anthracene C25H44 344.6 0.9530 0.9432 7.91 52.1 1.5081 1.036 0.909
Cholestane C27H48 372.8 0.9578 0.9482 21.15 103.4 1.52 a 1.050 0.897
1,1-Di(alphadecalyl)hendecane C31H56 428.8 0.9451 0.9356 20.25 99.5 1.5062 1.038 0.885
Aromatics
1,1-Diphenylethylene C14H12 180.2 1.0330 1.0235 3.96 b 39.1 b 1.6078 1.097 1.033
4,5-Dimethyl-9,10-dihydropenanthrene C16H14 206.3 1.1090 1.0998 4.93 42.2 1.6433 a 1.094 1.127
4,5-Dimethyl-9,10-dihydropenanthrene C16H16 208.3 1.0738 1.0684 4.43 40.8 1.6286 1.095 1.089
1,2-Diphenylbenze C18H18 230.3 1.0923 1.0814 4.53 41.2 1.6447 1.105 1.107
9-n-Butylanthracene C16H16 239.3 1.0600 1.0530 4.61 41.4 1.6006 1.074 1.064
1,2,3,4,5,6,8,13,14,15,15-Dodecahydrochrysene C18H24 240.4 1.0572 1.0492 8.16 52.9 1.5752 1.051 1.049
1,5-Di-phenyl-3(2-phenylethyl)2-pentene C25H28 326.5 1.0210 1.0159 3.42 37.6 1.5816 1.074 1.030
1,5-Di-phenyl-3(2-phenylethyl)pentane C25H40 328.5 1.0175 1.0076 3.81 38.8 1.5725 1.070 1.015
1,7-Dicyclopentyl-4-(2-phenylethyl)heptane C25H28 340.6 0.9251 0.9205 3.92 39.2 1.5070 t.047 0.919
1,2,3,4,5,6,8,9,10,17,18-Dodecahydro-9(n-octyl)naphthecene C26H40 352.6 0.9845 0.9796 13.09 70.5 1.5467 1.060 0.950
1,10-Di-(alphanaphyl)decane C30H34 394.6 1.0400 1.0296 12.27 67.4 1.6094 1.095 1.016
M: molecular weight;
SG: specific gravity at 15.56~ (60/60~
d: density at 20~ g/cm3;
v: kinematic viscosity at 210~ (99~ cSt;
V: Saybolt viscosity at 210~ (99~ SUS;
n: refractive index at 20~
Ra: refractivity intercept;
VGC: viscosity gravity constant. Data source: API-RP 42 [ 18].
a Calculated values of n.
bKinematic or Saybolt viscosities at 100~ (38~
 2. C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P U R E H Y D R O C A R B O N S
are calculated by Eqs. (2.14) and (2.15) and will be used to
develop predictive methods for the composition of heavy frac-
tions discussed in Chapter 3 (Section 3.5).
2.3 CHARACTERIZATION
OF HYDROCARBONS
The work on characterization of pure hydrocarbons began in
1933 when Watson and Nelson for the first time developed
two empirical charts relating molecular weight to either boil-
ing point and Kw or boiling point and API gravity [24]. In
these charts boiling point and specific gravity (or API grav-
ity) are used as the two independent input parameters. Since
then the work on characterization and methods of estimation
of basic properties of pure hydrocarbons and petroleum frac-
tions has continued to the present time. Methods developed
in 1930s till 1960s were mainly graphical, while with the use
of computer, methods developed in 1970s till present time are
in the forms of analytical correlations. The best example of
chart-type correlations, which has been in use by the industry,
is the Winn nomogram that relates molecular weight, critical
pressure, aniline point, and CH weight ratio to boiling point
and specific gravity [25]. A version of Winn nomograph as
used by the API [2] is presented in Fig. 2.12. Some of the an-
alytical correlations that are used in the industry are Cavett
[26], Kesler-Lee [12], Lee-Kesler [27], Riazi-Daubert [28, 29],
Twu [30], and Riazi-Sahhaf [31]. Most of these correlations
use boiling point and specific gravity as the input parame-
ters to estimate parameters such as molecular weight, critical
constants, and acentric factor. Most recently Korsten [32] has
developed a characterization scheme that uses boiling point
and a parameter called double-bond equivalent (DBE) as the
input parameters. DBE can be estimated from H/C atomic
ratio. In another paper Korsten [33] lists and evaluates vari-
ous correlations for estimation of critical properties of pure
hydrocarbons. Tsonopolus et al. [34] give the list of correla-
tions developed for characterization of coal liquids in terms
of boiling point and specific gravity. There are some meth-
ods of estimation of properties of pure compounds that are
based on various group contribution techniques. The most
accurate methods of group contributon for various proper-
ties with necessary recommendations are given in the fifth
edition of Properties of Gases and Liquids [4]. Even some of
these group contribution methods require properties such as
molecular weight or boiling point. Examples of such proce-
dures are the Lydersen and Ambrose methods [4]. The prob-
lem with group contribution methods is that the structure of
the compound must be known. For this reason they are not
appropriate for undefined petroleum fractions. However, they
can be used to predict properties of pure compounds when
experimental data are not available (i.e., critical properties
of heavy pure hydrocarbons). In fact, on this basis the prop-
erties of hydrocarbons heavier than C18 have been predicted
and reported by the API [2].
As discussed, during the past 70 years m a n y methods in
the forms of charts and equations were proposed to esti-
mate the basic properties of hydrocarbons from the knowl-
edge of the boiling point and specific gravity or the molecular
weight. Nearly all of these correlations are empirical in nature
45
without any theoretical explanation. Boiling point and spe-
cific gravity were used in most correlations based on expe-
rience and their availability. However, the characterization
methods proposed by Riazi and Daubert [28, 29, 35] are based
on the theory of intermolecular forces and EOS parameters
[36]. Although EOS are discussed in Chapter 5, their applica-
tion in the development of analytical correlations to charac-
terize hydrocarbons are discussed here. In the following parts
in this section several characterization schemes developed by
Riazi et al. [28, 29, 31, 35, 37] are presented along with other
methods.
2.3.1 Development of a Generalized Correlation
for Hydrocarbon Properties
Properties of a fluid depend on the intermolecular forces that
exist between molecules of that fluid [38, 39]. As summarized
by Prausnitz et al. [39] these forces are grouped into four cat-
egories. (1) Electrostatic forces between charged molecules
(ions) and between permanent dipoles or higher multipoles.
These forces result from the chemical structure of molecules
and are important in polar compounds (i.e., water, methanol,
ethanol, etc.). (2) Induction forces on molecules that are po-
larizable when subjected to an electric field from polar com-
pounds. These forces are also called dipole forces and are de-
termined by dipole moment of molecules 0z), which is propor-
tional to polarizability factor, ~, and the field strength. These
forces are proportional to/z 2 x ~. (3) The third type of forces
are attraction (dispersion forces) and repulsion between non-
polar molecules. These forces, also called London forces, are
static in nature and are proportional to u2. (4) The last are spe-
cial (chemical) forces leading to association or complex for-
mation such as chemical bonds. According to London these
forces are additive and except for very polar compounds, the
strongest forces are of the London type. For light and medium
hydrocarbon compounds, London forces are the dominant
force between the molecules.
The intermolecular force, F, is related to the potential en-
ergy, F, according to the following relation:
dF
(2.19) F --
dr
where r is the distance between molecules. The negative of
the potential energy, -F(r), is the work required to sepa-
rate two molecules from the intermolecular distance r to in-
finite separation. Equation of state parameters can be esti-
mated from the knowledge of the potential energy relation
[39]. Most hydrocarbon compounds, especially the light and
medium molecular weight hydrocarbons, are considered as
nonpolar substances. There are two forces of attraction (dis-
persion forces) and repulsion between nonpolar molecules.
The common convention is that the force of attraction is neg-
ative and that of repulsion is positive. As an example, when
molecules of methane are 1 nm apart, the force of attraction
between them is 2 x 10 -s dyne [39]. The following relation
was first proposed by Mie for the potential energy of nonpolar
molecules [39]:
Ao /3o
(2.20) F-
r n r m
46 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

where
No
(2.27) k OV2} =0
Prep --
- - -r n-
Application of Eqs. (2.26) and (2.27) to any two-parameter
Bo
Patt = --- EOS would result in relations for calculation of parameters
r m
A and B in terms of Tc and Pc, as shown in Chapter 5. It should
be noted that EOS parameters are generally designated by
Prep = repulsive potential lower case a and b, but here they are shown by A and B. No-
Fatt= attractive potential
tation a, b, c . . . . are used for correlation parameters in various
Ao = parameter characterizing the repulsive force (>0)
equations in this chapter. Applying Eqs. (2.26) and (2.27) to
/3o = parameter characterizing the attractive force (>0)
Eq. (2.25) results into the following three relations for To, Pc,
r --- distance between molecules
and Vc:
n, m = positive numbers, n > m
The main characteristics of a two-parameter potential energy (2.28) Tc = f4(A, B)
function is the m i n i m u m value of potential energy, P m i n , desig-
(2.29) Pc = fs(A, B)
nated by e = - Pmin and the distance between molecules where
the potential energy is zero (F = 0) which is designated by a. (2.30) Vc = f6(A, B)
L o n d o n studied the theory of dispersion (attraction) forces
Functions f4, f5, and f6 are universal functions and are the
and has shown that m = 6 and it is frequently convenient for
same for all fluids that obey the potential energy relation ex-
mathematical calculations to let n = 12. It can then be shown
pressed by Eq. (2.20) or Eq. (2.21). In fact, if parameters A
that Eq. (2.20) reduces to the following relation k n o w n as
and B in a two-parameter EOS in terms of Tc and Pc are rear-
Lennard-Jones potential [39]:
ranged one can obtain relations for Tc and Pc in terms of these
(2.21) P = 48 [ ( ~ ) 1 2 - ( ~ ) 6 1 two parameters. For example, for van der Waals and Redlich-
Kwong EOS the two parameters A and B are given in terms
of Tc and Pc [21] as shown in Chapter 5. By rearrangement of
In the above relation, e is a parameter representing molec-
the vdW EOS parameters we get
ular energy and a is a parameter representing molecular
size. Further discussion on intermolecular forces is given in
Section 5.3. To= AB-1 Pc= A B -2 Vc = 3B
According to the principle of statistical thermodynamics
there exists a universal EOS that is valid for all fluids that and for the Redlich-Kwong EOS we have
follow a two-parameter potential energy relation such as
Eq. (2.21) [40]. [ (0"0867)5R] 2/3 A2/3B-2/3
rC=k~ J
(2.22) Z = fl(g, a, T, P)
[ (0.0867) 5R 1 '/3 A 2/3B -s/3
Pc = L ~ J vc = 3.847B
(2.23) Z- PVr,~
RT
Similar relations can be obtained for the parameters of other
where
two-parameter EOS. A generalization can be made for the
Z = dimensionless compressibility factor
relations between To, Pc, and V~ in terms of EOS parameters
Vr, e = molar volume at absolute temperature, T, and pres-
A and B in the following form:
sure, P
/'1 = universal function same for all fluids that follow (2.31) [Tc, Pc, Vc] = aAbB c
Eq. (2.21).
By combining Eqs. (2.20)-(2.23) we obtain where parameters a, b, and c are the constants which differ
for relations for To, Pc, and Vc. However, these constants are
(2.24) Vr, v = f2(Ao, ]3o, T, P) the same for each critical property for all fluids that follow
where Ao and/3o are the two parameters in the potential en- the same two-parameter potential energy relation. In a two-
ergy relation, which differ from one fluid to another. Equation parameter EOS such as vdW or RK, Vc is related to only one
(2.24) is called a two-parameter EOS. Earlier EOS such as van parameter B so that Vc/B is a constant for all compounds.
der Waals (vdW) and Redlich-Kwong (RK) developed for sim- However, formulation of Vc t h r o u g h Eq. (2.30) shows that
ple fluids all have two parameters A and B [4] as discussed in Vc must be a function of two parameters A and B. This is
Chapter 5. Therefore, Eq. (2.24) can also be written in terms one of the reasons that two-parameters EOS are not accurate
of these two parameters: near the critical region. Further discussion on EOS is given
in Chapter 5.
(2.25) Vr, p = f3(A, B, T, P) To find the nature of these two characterizing parameters
The three functions fl, f2, and f3 in the above equations vary one should realize that A and B in Eq. (2.31 ) represent the two
in the form and style. The conditions at the critical point for parameters in the potential energy relation, such as e and c~ in
any PVT relation are [41 ] Eq. (2.22). These parameters represent energy and size char-
acteristics of molecules. The two parameters that are readily
(2.26) (0~) =0 measurable for h y d r o c a r b o n systems are the boiling point,
re.Pc Tb, and specific gravity, SG; in fact, Tb represents the energy
 2. C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF PURE H Y D R O C A R B O N S
parameter and SG represents the size parameter. Therefore,
in Eq. (2.31) one can replace parameters A and B by Tb and
SG. However, it should be noted that Tb is not the same as
parameter A and SG is not the same as parameter B, but it
is their combination that can be replaced. There are many
other parameters that may represent A and B in Eq. (2.31).
For example, if Eq. (2.25) is applied at a reference state of To
and P0, it can be written as
(2.32) Vr0,P0 = f3(A, B, To, Po)
where Vro,eo is the molar volume of the fluid at the reference
state. The most convenient reference conditions are tempera-
ture of 20~ and pressure of 1 atm. By rearranging Eq. (2.32)
one can easily see that one of the parameters A or B can be
molar volume at 20~ and pressure of 1 atm [28, 36, 42].
To find another characterization parameter we may con-
sider that for nonpolar compounds the only attractive force
is the London dispersion force and it is characterized by fac-
tor polarizability, a, defined as [38, 39]
(2.33) ct = ~ x x ~,r/2 q- 2 /
where
NA = Avogadro's number
M = molecular weight
p = absolute density
n = refractive index
In fact, polarizability is proportional to molar refraction, Rm,
defined as
(2.34) Rrn----- (-~-) X {n2-l'~,tz
2 q- 2`/
M
(2.35) V = --
p
n2-1
(2.36) I = - -
n2+2
in which V is the molar volume and I is a characterization
parameter that was first used by Huang to correlate hydrocar-
bon properties in this way [ 10, 42]. By combining Eqs. (2.34)-
(2.36) we get
Rm actual molar volume of molecules
(2.37) I- - Based on reported data in the 1977 edition of API-TDB, con-
V apparent molar volume of molecules
Rm, the molar refraction, represents the actual molar volume
of molecules, V represents the apparent molar volume and
their ratio, and parameter I represents the fraction of total
volume occupied by molecules. Rm has the unit of molar vol-
ume and I is a dimensionless parameter. Rm/M is the specific
refraction and has the same unit as specific volume. Parame-
ter I is proportional to the volume occupied by the molecules
and it is close to unity for gases (Ig ~ 0), while for liquids it
is greater than zero but less than 1 (0 < /liq <Z 1). Parameter
I can represent molecular size, but the molar volume, V, is a
parameter that characterizes the energy associated with the
molecules. In fact as the molecular energy increases so does
the molar volume. Therefore, both V and I can be used as two
independent parameters to characterize hydrocarbon proper-
ties. Further use of molar refraction and its relation with EOS
47
parameters and transport properties are discussed in Chap-
ters 5, 6, and 8. It is shown by various investigators that the
ratio of Tb/Tc is a characteristic of each substance, which is
related to either Tr or Tb [36, 43]. This ratio will be used to
correlate properties of pure hydrocarbons in Section 2.3.3.
Equation (2.31) can be written once for Tr in terms of V and I
and once for parameter Tb/Tr Upon elimination of parameter
V between these two relations, a correlation can be obtained
to estimate T~ from Tb and I. Similarly through elimination
of Tc between the two relations, a correlation can be derived
to estimate V in terms of Tb and I [42].
It should be noted that although both density and refrac-
tive index are functions of temperature, both theory and ex-
periment have shown that the molar refraction (Rm = VI) is
nearly independent of temperature, especially over a narrow
range of temperature [38]. Since V at the reference tempera-
ture of 20~ and pressure of 1 atm is one of the characteriza-
tion parameters, I at 20~ and 1 arm must be the other char-
acterization parameter. We chose the reference state of 20~
and pressure of 1 atm because of availability of data. Simi-
larly, any reference temperature, e.g. 25~ at which data are
available can be used for this purpose. Liquid density and re-
fractive index of hydrocarbons at 20~ and 1 atm are indicated
by d20 and n20, respectively, where for simplicity the subscript
20 is dropped in most cases. Further discussion on refractive
index and its methods of estimation are given elsewhere [35].
From this analysis it is clear that parameter I can be used
as one of the parameters A or B in Eq. (2.31) to represent the
size parameter, while Tb may be used to represent the energy
parameter. Other characterization parameters are discussed
in Section 2.3.2. In terms of boiling point and specific gravity,
Eq. (2.31) can be generalized as following:
(2.38) 0 = aT~SG c
where Tb is the normal boiling point in absolute de-
grees (kelvin or rankine) and SG is the specific gravity at
60 ~F(15.5 ~C). Parameter 0 is a characteristic property such as
molecular weight, M, critical temperature, To, critical pres-
sure, Pc, critical molar volume, Vc, liquid density at 20~
d20, liquid molar volume at 20~ and I atm, V20, or refrac-
tive index parameter, I, at 20~ It should be noted that
0 must be a temperature-independent property. As mentioned
before, I at 20~ and 1 atm is considered as a character-
istic parameter and not a temperature-dependent property.
stants a, b, and c were determined for different properties
and have been reported by Riazi and Daubert [28]. The con-
stants were obtained through linear regression of the loga-
rithmic form of Eq. (2.38). Equation (2.38) in its numerical
form is presented in Sections 2.4-2.6 for basic characteri-
zation parameters. In other chapters, the form of Eq. (2.38)
will be used to estimate the heat of vaporization and trans-
port properties as well as interconversion of various distil-
lation curves. The form of Eq. (2.38) for T~ is the same as
the form Nokay [44] and Spencer and Daubert [45] used to
correlate the critical temperature of some hydrocarbon com-
pounds. Equation (2.38) or its modified versions (Eq. 2.42),
especially for the critical properties and molecular weight,
have been in use by industry for many years [2, 8, 34, 46-56].
Further application of this equation will be discussed in
Section 2.9.
48 CHARACTERIZATION AND P R O P E R T I E S OF P E T R O L E U M FRACTIONS
One should realize that Eq. (2.38) was developed based on
a two-parameter potential energy relation applicable to non-
polar compounds. For this reason, this equation cannot be
used for systems containing polar compounds such as alco-
hol, water, or even some complex aromatic compounds that
are considered polar. In fact constants a, b, and c given in
Eq. (2.38) were obtained based on properties of hydrocar-
bons with carbon number ranging from Cs to C20. This is
almost equivalent to the molecular weight range of 70-300.
In fact molecular weight of n-C20 is 282, but considering the
extrapolation power of Eq. (2.38) one can use this equation
up to molecular weight of 300, which is roughly equivalent to
boiling point of 370~ (700~ Moreover, experimental data
on the critical properties of hydrocarbons above C~8 were not
available at the time of development of Eq. (2.38). For heav-
ier hydrocarbons additional parameters are required as will
be shown in Section 2.3.3 and Chapter 3. The lower limit for
the hydrocarbon range is C5, because lighter compounds of
C1-C4 are mainly paraffinic and in the gaseous phase at nor-
mal conditions. Equation (2.38) is mainly applied for unde-
fined petroleum fractions that have average boiling points
higher than the boiling point of C5 as will be seen in Chapter
3. Methods of calculation of properties of natural gases are
discussed in Chapter 4.
Example 2.3--Show that the molecular weight of hydrocar-
bons, M, can be correlated with the boiling point, Tb, and
specific gravity, SG, in the form of Eq. (2.38).
Solution--It was already shown that molar volume at 20~
and l atm, V20, can be correlated to parameters A and B of a
potential energy function through Eq. (2.32) as follows: V20 =
gl(A, B). In fact, parameter V2o is similar in nature to the
critical molar volume, Vr and can be correlated with Tb and
SG as Eq. (2.38): V20 = aTbbSGc. But V20 = M/d2o, where d20
is the liquid density at 20~ and 1 atm and is considered a
size parameter. Since Tb is chosen as an energy parameter
and SG is selected as a size parameter, then both d20 and SG
represent the same parameter and can be combined with the
energy parameter as M = aT~SG r which has the same form
of Eq. (2.38) for 0 = M.
2.3.2 Various Characterization Parameters
for Hydrocarbon Systems
Riazi and Daubert [29] did further study on the expansion
of the application of Eq. (2.38) by considering various in-
put parameters. In fact instead of Tb and SG we may con-
sider any pair of parameters 0a and 02, which are capable of
TABLE 2.4--Comparison of properties of adjacent members of paraffin family and two families
of C8 hydrocarbons.
Hydrocarbongroup Tb,K M v38000),cSt
Paraffin family
C7Hi6(n-heptane) 371.6 100.2
C8H18(n-octane) 398.8 114.2
% Difference in property +7.3 +14.2
Two Families (C8)
Paraffin (n-octane) 398.8 114.2
Aromatic (ethylbenzene) 409.4 106.2
% Difference in property +2.7 -7.0
characterizing molecular energy and the molecular size. This
means that Eq. (2.31) can be expressed in terms of two pa-
rameters 01 and 02:
(2.39) 0 = aOblO~
However, one should realize that while these two parame-
ters are independent, they should represent molecular energy
and molecular size. For example, the pairs such as (Tb, M)
or (SG, I) cannot be used as the pair of input parameters
(01, 02). Both SG and I represent size characteristics of
molecules and they are not a suitable characterization pair.
In the development of a correlation to estimate the proper-
ties of hydrocarbons, all compounds from various hydrocar-
bon groups are considered. Properties of hydrocarbons vary
from one hydrocarbon type to another and from one carbon
number to another. Hydrocarbons and their properties can
be tabulated as a matrix of four columns with many rows.
Columns represent hydrocarbon families (paraffins, olefins,
naphthenes, aromatics) while rows represent carbon num-
ber. Parameters such as Tb, M, or kinematic viscosity at 38~
(100~ v38(100), vary in the vertical direction with carbon
number, while parameters such as SG, I, and CH vary signif-
icantly with hydrocarbon type. This analysis is clearly shown
in Table 2.4 for C8 in paraffin and aromatic groups and prop-
erties of C7 and C8 within the same group of paraffin family.
As is clearly shown by relative changes in various properties,
parameters SG, I, and CH clearly characterize hydrocarbon
type, while Tb, M, and 1)38(100) are good parameters to charac-
terize the carbon number within the same family. Therefore,
a correlating pair should be selected in a way that character-
izes both the hydrocarbon group and the compound carbon
number. A list of properties that may be used as pairs of cor-
relating parameters (01,02) in Eq. (2.39) are given below [29].
Tc (K), Pc (bar), Vc (cm3/g), M, Tb (K), SG, I (20~ CH
(Tb, SG), (Tb, I), (Tb, CH), (M, SG), (M, I), (M, CH),
(01, 02)
Pairs: (v38(100), SG), (v38(100),I), (v38(100), CH)
The accuracy of Eq. (2.39) was improved by modification of
its a parameter in the following form:
(2.40) 0 = a exp[b0a + c02 "[-dOlO2]O~Of2
Values of constants a - f in Eq. (2.40) for various param-
eters of 0 and pairs of (01, Oz) listed above are given in
Table 2.5. It should be noted that the constants reported by
Riazi-Daubert have a follow-up correction that was reported
later in the same volume [29]. These constants are obtained
from properties of hydrocarbons in the range of C5-C20 in
SG 1 CH
0.5214 0.6882 0.236 5.21
0.6476 0.7070 0.241 5.30
+24.2 +2.7 +2.1 +1.7
0.6476 0.7070 0.241 5.30
0.6828 0.8744 0.292 9.53
5.4 +23.7 +21.2 +79.8
 T A B L E 2.5--Coefficients o f Eq. (2.40) 0 = a exp (bO1 + c02 + dO102)O~Of for various properties.
0 01 02 a b c d e f AAD%
Tc Tb SG 9.5232 - 9 . 3 1 4 • 10 -4 -0.54444 6.48 x 10 -4 0.81067 0.53691 0.5
Tc Tb I 4.487601 • 105 - 1 . 3 1 7 1 • 10 -3 -16.9097 4.5236 • 10 -3 0.6154 4.3469 0.6
Tc Tb CH 8.6649 0 0 0 0.67221 0.10199 0.7
Tc M SG 308 - 1 . 3 4 7 8 • 10 -4 -0.61641 0 0.2998 1.0555 0.7
Tc M I 1.347444 • 106 2.001 • 10 -4 -13.049 0 0.2383 4.0642 0.8
Tc M CH 20.74 1.385 • 10 -3 -0.1379 -2.7 • 10 -4 0.3526 1.4191 1.0
Tc v38(100) I 2.4522 • 103 --0.0291 -- 1.2664 0 0.1884 0.7492 1.0
Pc Tb SG 3.195846 x 105 - 8 . 5 0 5 • 10 -3 -4.8014 5.749 x 10 -3 -0.4844 4.0846 2.7
Pc Tb I 8.4027 • 1023 --0.012067 --74.5612 0.0342 --1.0303 18.4302 2.6
Pc Tb CH 9.858968 - 3 . 8 4 4 3 x 10 -3 -0.3454 0 -0.1801 3.2223 3.5
Pc M SG 3116.632 - 1 . 8 0 7 8 • 10 -3 -0.3084 0 -0.8063 1.6015 2.7
Pc M I 2.025 x 1016 -0.01415 -48.5809 0.0451 -0.8097 12.9148 2.3
Pc M CH 56.26043 - 2 . 1 3 9 x 10 -3 -0.265 0 -0.6616 2.4004 3.1
Pc v38(lO0) I 393.306 0 0 0 -0.4974 2.052 4.2
Vc Tb SG 6.049 • 10 -2 - 2 . 6 4 2 2 x 10 3 -0.26404 1.971 • 10 -3 0.7506 -1.2028 1,8
Vc Tb I 6.712 • 10 -3 --2.720 • 10 -3 0.91548 7.920 x 10 -3 0.5775 --2.1548 1.8
Vc Tb CH 1.409 • 10 +1 - 1 . 6 5 9 4 x 10 -3 0.05345 2.6649 x 10 -4 0.1657 -1.4439 2.4
Vc M SG 7.529 x 10 -1 - 2 . 6 5 7 x 10 -3 0.5287 2.6012 • 10 -3 0.20378 -1,3036 1.9
Vc M I 6.3429 • 10 -5 - 2 . 0 2 0 8 • 10 -3 14.1853 4.5318 x 10 -3 0.2556 -4.60413 2.1
Vc M CH 1.597 x 10 +1 - 2 . 3 5 3 3 • 10 -3 0.1082 3.826 x 10 -4 0.0706 -1.3362 2.6
Vc v38(lO0) I 2.01 x 10 -1~ -0.16318 36.09011 0.4608 0.1417 - 10.65067 1.7
M Tb SG 1032.1 9.78 • 10 -4 -9.53384 2.0 • 10 -3 0 .9 7 4 7 6 6.51274 2.1
M Tb I 8.9205 x 10 -6 1.55833 x 10 -5 4.2376 0 2.0935 -1.9985 2.3
M Tb CH 1.81456 x 10 -3 0 0 0 1.9273 -0.2727 2.3
M v38(too) SG 1.51723 • 106 -0.195411 -9.638970 0.162470 0 .5 6 3 7 0 6.89383 3.4
M v38(lOO) I 4 x 10 -9 - 8 . 9 8 5 4 x 10 -2 38.106 0 0.6675 -10.6 3.5
Tb M SG 3.76587 3.77409 x 10 -3 2.984036 - 4 . 2 5 2 8 8 • 10 -3 4.0167 x 10 -1 -1.58262 1.0
Tb M I 75.775 0 0 0 0.4748 0.4283 1.1
Tb M CH 2 0. 253 4 7 - 1 . 5 7 4 1 5 x 10 -4 - 4 . 5 7 0 7 • 10 -2 9.22926 • 10 -6 0 .5 1 2 9 7 6 0.472372 1.1
Tb v38(i00) SG 237 9 . 86 - 1 . 3 0 5 1 • 10 -2 -1.68759 - 2 . 1 2 4 7 x 10 -2 0 .2 6 2 9 1 4 1.3489 1.7
Tb v38(lO0) I 0 . 0 506 2 9 - 6 . 5 2 3 6 x 10 -2 14.9371 6.029 x 10 -2 0.3228 -3.8798 1.6
SG Tb I 2.43810 • 107 -4.19400 • 10 -4 -23.553500 3.98736 • 10 -3 -0.341800 6.9195 0.5
SG Tb CH 2.86706 x 10 -3 -1.83321 • 10 -3 -0.081635 6.49168 x 10 -5 0.890041 0.73238 1.0
SG M I 1.1284 x 106 -1.58800 x 10 -3 -20.594000 7.34400 x 10 -3 -0.07710 6.30280 0.5
SG M CH 6.84403 • 10 -2 -1,48844 • 10 -3 -0.079250 4.92112 x 10 -5 0 .2 8 9 8 4 4 0.919255 1.3
SG v38(lOO) I 3.80830 • 107 -6.14060 x 10 -2 -26.393400 0.253300 -0.235300 8.04224 0.5
I Tb SG 0.02343 7.0294 • 10 -4 2.46832 -1.0268 x 10 -3 0.05721 -0.71990 0.5
I Tb CH 5.60121 • 10 -3 - 1 . 7 7 7 4 • 10 -4 - 6 . 0 7 3 7 x 10 -2 -7.9452 • 10 -5 0.44700 0 .9 8 9 6 0 0.5
I M SG 0.42238 3.1886 • 10 -4 -0.200996 -4.2451 x 10 -4 -0.00843 1.11782 0.5
I M CH 4.23900 x 10 2 - 5 . 6 9 4 6 • 10 -4 - 6 . 8 3 6 x 10 -2 0.0 0 .1 6 5 6 0 0,82910 0.9
I v38(loo SG 0. 26 37 60 1.7458 x 10 -2 0.231043 -1.8441 • 10 -2 -0.01128 0.77078 0.6
I v38(lOO CH 8.71600 • 10 -2 6.1396 • 10 -3 - 7 . 0 1 9 • 10 -2 -2.5935 • 10 -3 0.05166 0.84599 0.4
CH Tb SG 3.47028 1.4850 • 10 -2 16.94020 -0.01.2491 -2.72522 -6.79769 2.5
CH Tb I 8.39640 • 10 -13 7.7171 • 10 -3 71.65310 --0.02.088 --1.37730 --13.61390 1.7
CH M SG 2.35475 9.3485 x 10 -3 4.74695 - 8 . 0 1 7 2 • 10 -3 -0.68418 -0.76820 2.6
CH M I 2.9004 • 10 -23 7.8276 x 10 -3 60.34840 - 2 . 4 4 5 0 • 10 -2 -0.37884 -12.34051 1.7
CH v38(100 SG 2.52300 • 10-12 0.48281 29.98797 -0.55768 -0.14657 -20.31303 2.5
CH v38(lO0 I 2.1430 • 10 -12 0.28320 53.73160 -0.91085 -0.17158 -10.88065 1.6
Coefficients for Eq. (2.40) have been obtained from data on approximately 140 pure hydrocarbons in the molecular weight range of 70-300 and boiling point range of 300-620 K. Units: Vc in cma/g, Tb and
Tc in K, Pc in bar, and v38(100)in cSt. With permission from Ref. [29].
50 CHARACTERIZATION AND PROPERTIES OF PETROLEUM
a similar approach as the constants of Eq. (2.38) were ob-
tained. Similarly, Eq. (2.40) can be applied to hydrocarbon
systems in the molecular weight range of 70-280, which is
approximately equivalent to the boiling range of 30-350~
("-80-650~ However, they may be used up to C22 o r molec-
ular weight of 300 (~boiling point 370~ with good accuracy.
In obtaining the constants, Eq. (2.40) was first converted into
linear form by taking logarithm from both side of the equation
and then using a linear regression program in a spreadsheet
the constants were determined. The value of R 2 (index of cor-
relation) is generally above 0.99 and in some cases near 0.999.
However, when viscosity or CH parameters are used the R 2
values are lower. For this reason use of kinematic viscosity or
CH weight ratio should be used as a last option when other pa-
rameters are not available. Properties of heavy hydrocarbons
are discussed in the next section. When Eq. (2.40) is applied
to petroleum fractions, the choice of input parameters is de-
termined by the availability of experimental data; however,
when a choice exists the following trends determine the char-
acterizing power of input parameters used in Eq. (2.39) or
(2.40): The first choice for 01 is Tb, followed by M, and then
v38(100), while for the parameter 02 the first choice is SG, fol-
lowed by parameters I and CH. Therefore the pair of (M, SG)
is preferable to (M, CH) when the choice exists.
2.3.3 Prediction of Properties o f H e a v y
Pure Hydrocarbons
One of the major problems in characterization of heavy
petroleum fractions is the lack of sufficient methods to predict
basic characteristics of heavy hydrocarbons. As mentioned in
the previous section, Eqs. (2.38) or (2.40) can be applied to
hydrocarbons up to molecular weight of about 300. Crude oils
and reservoir fluids contain fractions with molecular weights
higher than this limit. For example, products from vacuum
distillation have molecular weight above this range. For such
fractions application of either Eq. (2.38) or (2.40) leads to
some errors that will affect the overall property of the whole
crude or fluid. While similar correlations may be developed
for higher molecular weight systems, experimental data are
limited and most data (especially for critical properties for
such compounds) are predicted values. As mentioned in the
previous section, the heavy hydrocarbons are more complex
and two parameters may not be sufficient to correlate prop-
erties of these compounds.
One way to characterize heavy fractions, as is discussed in
the next chapter, is to model the fraction as a mixture of pseu-
docompounds from various homologous hydrocarbon fami-
lies. In fact, within a single homologous hydrocarbon group,
such as n-alkanes, only one characterization parameter is suf-
ficient to correlate the properties. This single characterization
parameter should be one of those parameters that best char-
acterizes properties in the vertical direction such as carbon
number (Nc), Tb, or M. As shown in Table 2.4, parameters
SG, I20, and CH weight ratio are not suitable for this purpose.
Kreglewiski and Zwolinski [57] used the following relation to
correlate critical temperature of n-alkanes:
(2.41) ln(0~ - O) = a - bN2c/3
where 0~ represents value of a property such as Tc at Nc -~ ~ ,
and 0 is the value of Tc for the n-alkane with carbon number of
FRACTIONS
Nc. Later, this type of correlation was used by other investiga-
tors to correlate Tc and Pc for n-alkanes and alkanols [58-60].
Based on the above discussion, M or Tb may also be used in-
stead of Nc. Equation (2.41) suggests that for extremely high
molecular weight hydrocarbons (M -+ or critical tempera-
ture or pressure approaches a finite value (Tcoo, Pc~). While
there is no proof of the validity of this claim, the above equa-
tion shows a good capability for correlating properties of
n-alkanes for the molecular weight range of interest in prac-
tical applications.
Based on Eq. (2.41), the following generalized correlation
was used to characterize hydrocarbons within each homolo-
gous hydrocarbon group:
(2.42) ln(0~ - O) = a - b M c
The reason for the use of molecular weight was its avail-
ability for heavy fractions in which boiling point data may
not be available due to thermal cracking. For four groups
of n-alkanes, n-alkylcycopentanes, n-alkylcyclohexanes, and
n-alkylbenzenes, constants in Eq. (2.42) were determined us-
ing experimental data reported in the 1988 edition of API-TDB
[2] and 1986 edition of TRC [21]. The constants for TM, Tb, SG,
dE0, I , Tbr (Tb/Tc), Pc, dc, w, and cr are given in Table 2.6 [31].
Carbon number range and absolute and average absolute de-
viations (AAD) for each property are also given in Table 2.6.
Errors are generally low and within the accuracy of the ex-
perimental data. Equation (2.42) can be easily reversed to
estimate M from Tb for different families if Tb is chosen as the
characterizing parameter. Then estimated M from Tb can be
used to predict other properties within the same group (fam-
ily), as is shown later in this chapter. Similarly if Nc is chosen
as the characterization parameter, M for each family can be
estimated from Nc before using Eq. (2.42) to estimate various
properties. Application and definition of surface tension are
discussed in Chapter 8 (Sec 8.6).
Constants given in Table 2.6 have been obtained from the
properties of pure hydrocarbons in the carbon number ranges
specified. For TM, Tb, SG, d, and I, properties of compounds
up to C40 were available, but for the critical properties values
up to C20 were used to obtain the numerical constants. One
condition imposed in obtaining the constants of Eq. (2.42)
for the critical properties was the criteria of internal con-
sistency at atmospheric pressure. For light compounds crit-
ical temperature is greater than the boiling point (Tbr < 1)
and the critical pressure is greater than 1 atm (Pc > 1.01325
bar). However, this trend changes for very heavy compounds
where the critical pressure approaches 1 atm. Actual data
for the critical properties of such compounds are not avail-
able. However, theory suggests that when Pc --~ 1.01325 bar,
Tc ~ Tb or Tb~ ~ 1. And for infinitely large hydrocarbons
when Nc --~ ~ (M --~ oo), Pc --~ 0. Some methods developed
for prediction of critical properties of hydrocarbons lead to
Tbr =- 1 as Nc -~ oo[43]. This can be true only if both Tc and
Tb approach infinity as Nc ~ oo. The value of carbon num-
ber for the compound whose Pc = 1 atm is designated by N~.
Equation (2.42) predicts values of Tb~ = 1 at N~ for different
homologous hydrocarbon groups. Values of N~ for different
hydrocarbon groups are given in Table 2.7. In practical ap-
plications, usually values of critical properties of hydrocar-
bons and fractions up to C45 or C50 are needed. However,
accurate prediction of critical properties at N~* ensures that
 2. C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P U R E H Y D R O C A R B O N S 51

TABLE 2.6---Constants of Eq. (2.42) for various parameters.

Constants in Eq. (2.42)
0 C N o . Range 00r a b c AADb %AADb
Constants for physical properties of n-alkanes [3 1]a
TM Cs-C40 397 6.5096 0.14187 0.470 1.5 0.71
Tb C5-C40 1070 6.98291 0.02013 2/3 0.23 0.04
SG C5-C19 0.85 92.22793 89.82301 0.01 0.0009 0.12
d20 C5-C40 0.859 88.01379 85.7446 0.01 0.0003 0.04
I C5-C4o 0.2833 87.6593 86.62167 0.01 0.00003 0.002
Tbr ----Tb/Tc C5-C20 1.15 -0.41966 0.02436 0.58 0.14 0.027
-Pc C5-C20 0 4.65757 0.13423 0.5 0.14 0.78
de C5-C20 0.26 -3.50532 1.5 x 10 -8 2.38 0.002 0.83
-to C5-C2o 0.3 -3.06826 -1.04987 0.2 0.008 1.2
a C5-C20 33.2 5.29577 0.61653 0.32 0.05 0.25
Constants for physical properties of n-alkylcyclopentanes
TM C7-C41 370 6.52504 0.04945 2/3 1.2 0.5
Tb C6-C41 1028 6.95649 0.02239 2/3 0.3 0.05
SG C7-C25 0.853 97.72532 95.73589 0.01 0.0001 0.02
d20 C5-C41 0.857 85.1824 83.65758 0.01 0.0003 0.04
I C5-C4I 0.283 87.55238 86.97556 0.01 0.00004 0.003
Tbr ----Tb/Tc C5-C18 1.2 0.06765 0.13763 0.35 1.7 0.25
-Pc C6-C18 0 7.25857 1.13139 0.26 0.4 0.9
-de C6-C20 -0.255 -3.18846 0.1658 0.5 0.0004 0.11
-to C6-C20 0.3 -8.25682 -5.33934 0.08 0.002 0.54
o" C6-C25 30.6 14.17595 7.02549 0.12 0.08 0.3
Constants for physical properties of n-alkylcyclohexane
TM C7-C20 360 6.55942 0.04681 0.7 1.3 0.7
Tb C6-C20 1100 7.00275 0.01977 2/3 1.2 0.29
SG C6-C20 0.845 -1.51518 0.05182 0.7 0.0014 0.07
d20 C6-C21 0.84 - 1.58489 0.05096 0.7 0.0005 0.07
I C6-C20 0.277 -2.45512 0.05636 0.7 0.0008 0.06
Tbr = Tbfrc C6-C20 1.032 -0.11095 0.1363 0.4 2 0.3
-Pc C6-C20 0 12.3107 5.53366 0.1 0.15 0.5
-dc C6-C20 -0.15 -1.86106 0.00662 0.8 0.0018 0.7
--CO C7-C20 0.6 -5.00861 -3.04868 0.1 0.005 1.4
a C6-C20 31 2.54826 0.00759 1.0 0.17 0,6
Constants for physical properties of n-alkylbenzenes
TM C9-C42 375 6.53599 0.04912 2/3 0.88 0.38
Tb C6-C42 1015 6.91062 0.02247 2/3 0.69 0.14
-SG C6-C20 -0.8562 224.7257 218.518 0.01 0.0008 0.1
-d20 C6-C42 -0.854 238.791 232.315 0.01 0.0003 0.037
-I C6-C42 -0.2829 137.0918 135.433 0.01 0.0001 0.008
Tbr = Tb/Tc C6-C20 1.03 -0.29875 0.06814 0.5 0.83 0.12
-Pc C6-C20 0 9.77968 3.07555 0.15 0.22 0.7
-de C6-C20 -0.22 -1.43083 0.12744 0.5 0.002 0.8
-co C6-C20 0 -14.97 -9.48345 0.08 0.003 0.68
tr C6-C20 30.4 1.98292 -0.0142 1.0 0.4 1.7
With permission from Ref. [31].
aData sources: TMTb, and d are taken from TRC [21]. All other properties are taken from API-TDB-1988 [2]. Units:
TM, Tb, and Tc are in K; d20 and d~ are in g/cm3; Pc is in bar; a is in dyn/cm.
b AD and AAD%given by Eqs. (2.134) and (2.135).

t h e e s t i m a t e d c r i t i c a l p r o p e r t i e s b y Eq. (2.42) a r e r e a l i s t i c for co, a n d a all h a v e finite values. F r o m a p h y s i c a l p o i n t o f

h y d r o c a r b o n s b e y o n d Cla. T h i s a n a l y s i s is c a l l e d internal con-
sistency for c o r r e l a t i o n s of c r i t i c a l p r o p e r t i e s .
I n t h e c h a r a c t e r i z a t i o n m e t h o d p r o p o s e d b y K o r s t e n [32,
33] it is a s s u m e d t h a t for e x t r e m e l y l a r g e h y d r o c a r b o n s
(Nc ~ o~), t h e b o i l i n g p o i n t a n d critical t e m p e r a t u r e also ap-
p r o a c h infinity. H o w e v e r , a c c o r d i n g to Eq. (2.42) as Nc --~ er
o r (M--~ o~), p r o p e r t i e s s u c h as Tb, SG, d, I, Tbr, Pc, de,
v i e w this m a y be t r u e for m o s t of t h e s e p r o p e r t i e s . H o w e v e r ,
K o r s t e n [33] s u g g e s t s t h a t as Nc --~ o~, Pc a n d dc a p p r o a c h
z e r o w h i l e Tb, To, a n d m o s t o t h e r p r o p e r t i e s a p p r o a c h infinity.
G o o s s e n [61] d e v e l o p e d a c o r r e l a t i o n for m o l e c u l a r w e i g h t of
h e a v y f r a c t i o n s t h a t s u g g e s t s b o i l i n g p o i n t for e x t r e m e l y l a r g e
m o l e c u l e s a p p r o c h e s a finite v a l u e of Tbo~ = 1078. I n a n o t h e r
p a p e r [62] he s h o w s t h a t for infinite p a r a f f i n i c c h a i n length,

TABLE 2.7--Prediction of atmospheric critical pressure from Eq. (2.42).

Nc*calculated at N* calculated at Predicted Pc (bar) at
Hydrocarbon type Tb = Tc Pc ~ 1.01325 Tb = Tc
n-Mkanes 84.4 85 1.036
n-Mkylcyclopentanes 90.1 90.1 1.01
n-Mtcylcyc]ohexanes 210.5 209.5 1.007
n-Mkylbenzenes 158.4 158.4 1,013
With permission from Ref. [31 ].
52 CHARACTERIZATION AND P R O P E R T I E S OF P E T R O L E U M FRACTIONS

1.2

~n-alkanes
1.1
. . . . n-all~,lcyclope~taaes
. . . . . ,-alkylbenzenes ~ / ~;~. . . _ .
1.0
b.

o 0.9

~ 0.8
o
o

o
0.7

./
0.6

5 L i i ~ i j J~l i i i I i i ill K i J , , J ,I

10 100 1000 10000

Molecular Weight

FIG. 2.1--Reduced boiling point of homologous hydrocarbon groups from

Eq. (2.42).

do~ = 0.8541 and no~ = 1.478 (Ioo = 0.283), while the values Pan et al. [63, 64] also r e c o m m e n d use of the following re-
obtained t h r o u g h Eq. (2.42) (see Table 2.6) are Tboo= 1070, lation for the acentric factor of aromatics for hydrocarbons
doo = 0.859, and Ioo = 0.2833. One can see h o w close the val- with M < 800:
ues are although they have been derived by two different
(2.44) In to = -36.1544 + 30.94M 0"026261
methods. However, these values are of little practical applica-
tion as long as a proposed correlation satisfies the condition and when M > 800, to = 2.0. Equation (2.42) is r e c o m m e n d e d
of Tbr = 1 at Pc = 1.0133 bar. Equation (2.42) will be used later for calculation of other t h e r m o d y n a m i c properties based on
in Chapter 4 to develop physical properties of single carbon the evaluation made on t h e r m o d y n a m i c properties of waxes
n u m b e r (SCN) cuts up to C50 for the estimation of properties and asphaltenes [63, 64].
of heavy crude oils and reservoir oils. Graphical presentation For homologous h y d r o c a r b o n groups, various correlations
of Eq. (2.42) for Tbr and Tc versus molecular weight of dif- m a y be found suitable for the critical properties. For example,
ferent hydrocarbon families is shown in Figs. 2.1 and 2.2 for another relation that was found to be applicable to critical
molecular weights up to 3000 (Nc ~ 214). pressure of n-alkyl families is in the following form:
One direct application of critical properties of homologous
hydrocarbons is to calculate phase equilibrium calculations (2.45) Pc = (a + bM)-"
for wax precipitation and cloud point of reservoir fluids and where Pc is in bar and M is the molecular weight of pure hy-
crude oils as shown by Pan et al. [63, 64]. These investigators drocarbon from a homologous group. Constant n is greater
evaluated properties calculated through Eq. (2.42) and mod- than unity and as a result as M ~ oo we have Pc ~ 0, which
ified this equation for the critical pressure of PNA hydrocar- satisfies the general criteria for a Pc correlation. Based on data
bons with molecular weight above 300 t h r o u g h the following on Pc of n-alkanes from C2 to C22, as given in Table 2.1, it was
relation: found that n = 1.25, a = 0.032688, and b = 0.000385, which
(2.43) Pc = a - b e x p ( - c M ) gives R 2 = 0.9995 with average deviation of 0.75% for 21 com-
pounds. To show the degree of extrapolation of this equation,
where a, b, and c are given for the three h y d r o c a r b o n groups if data from C2 to C10 (only nine compounds) are used to
in Table 2.8 [64]. However, Eq. (2.43) does not hold the inter-
nal consistency at Pc of I atm, which was imposed in deriving
the constants of Eq. (2.42). But this m a y not affect results for TABLE 2.8---Coefficients of Eq. (2.43).
practical calculations as critical pressures of even the heaviest Coefficient Paraffins Naphthenes Aromatics
a 0.679091 2.58854 4.85196
c o m p o u n d s do not reach to atmospheric pressure. A compar- b -22.1796 -27.6292 -42.9311
ison between Eq. (2.42) and (2.43) for the critical pressure c 0.00284174 0.00449506 0.00561927
of paraffins, naphthenes, and aromatics is shown in Fig. 2.3. Taken from PanetaL[63,64].
 2. C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P U R E H Y D R O C A R B O N S 53

700

* API-TDB
--n-ahkanes
. . . . . . . n-alkylcyclopentanes
. . . . n-alkylbenzenes

500
?

o)

#
-~ 300

100
lO 100 1000

Molecular Weight

FIG. 2.2--Critical temperature of homologous hydrocarbon groups from

Eq. (2,42).

50 o

",x
*" t . DIPPR Data
.~ R-S: Eq. 2.42
40 ., ~ - - n-alk~es

.:~ ....... n-alkyleyclopentanes

~ .... n-alkylb~zenes
~ ~ ~ P-F: Eq. (2.43)
m 30 --n-alkanes
o~
.... n-all~,lcyclopentanes
X - - - - n-alkylbeazenes

~ 20

10

I0 I00 1000 10000

Molecular Weight

FIG. 2.3--Prediction of critical pressure of homologous hydrocarbon groups from

Eqs. (2.42) and (2.43).
54 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

TABLE 2.9--Constants in Eqs. (2.46a and 2.46b) 0 = al exp(bl01 + ClSG + dl01SG) 0~SGf for various properties of heavy hydrocarbons.
o o1 al bi cl dl el fl AAD%
Tc Tb 35.9413 -6.9 • 10-4 -1.4442 4.91 • 10-4 0.7293 1.2771 0.3
Pc Tb 6.9575 -0.0135 -0.3129 9.174 • 10 3 0.6791 -0.6807 5.7
Vc Tb 6.1677 x 101~ -7.583 • 10-3 -28.5524 0.01172 1.20493 17.2074 2.5
I Tb 3.2709 • 10-3 8.4377 • 10-4 4.59487 -1.0617 • 10-3 0.03201 -2.34887 0.l
d20 Tb 0.997 2.9 • 10-4 5.0425 -3.1 • 10-4 -0.00929 1.01772 0.07
0 01 32 b2 c2 d2 e2 ~ AAD%
Tb M 9.3369 1.65 • 10-4 1.4103 -7.5152 x 10-4 0.5369 -0.7276 0.3
Tc M 218.9592 -3.4 • 10-4 -0.40852 -2.5 • 10-5 0.331 0.8136 0.2
Pc M 8.2365 • 104 -9.04 • 10-3 -3.3304 0.01006 -0.9366 3.1353 6.2
Vc M 9.703 • 106 -9.512 • 10-3 -15.8092 0.01111 1.08283 10.5118 1.6
I M 1.2419 x 10-2 7.27 • 10-4 3.3323 -8.87 x 10-4 6.438 • 10-3 -1.61166 0.2
d20 M 1.04908 2.9 • 10 4 -7.339 • 10-2 -3.4 • 10 - 4 3.484 • 10-3 1.05015 0.09
Data generatedfrom Eq. (2.42) have been used to obtain these constants. Units:Vcin cm3/mol;To,and Tbin K;
Pc in bar; d20 in g/cm3 at 20~ Equations are recommendedfor the carbon range of C20-C50;however, they may be used for the C5-C20with lesser degree of
accuracy.

o b t a i n a a n d b in the above e q u a t i o n we get a = 0.032795 a n d extended in terms of three parameters, Tb, d20, a n d M, to es-
b = 0.000381. These coefficients give R 2 ----0.9998 b u t w h e n it timate the critical properties of both h y d r o c a r b o n s a n d n o n -
is used to estimate Pc from C2 to C22 AAD of 0.9% is obtained. h y d r o c a r b o n s [37].
These coefficients estimate Pc of t/-C36 a s 6.45 b a r versus value
given in DIPPR as 6.8. This is a good extrapolation power. I n Tc, Pc, Vc = exp[a + bM + cTb + dd20 + eTbd2o] MfT~+hMdi2o
Eq. (2.45) one m a y replace M by Tb or Nc a n d o b t a i n n e w (2.47)
coefficients for cases that these parameters are known.
Based o n the critical properties of more t h a n 170 hydrocar-
Properties of pure c o m p o u n d s predicted t h r o u g h Eqs.
bons from C1 to C18 a n d more t h a n 80 n o n h y d r o c a r b o n s , such
(2.42) a n d (2.43) have b e e n used to develop the following gen-
as acids, sulfur c o m p o u n d s , nitriles, oxide gases, alcohols,
eralized correlations in terms of (Tb, SG) or (M, SG) for the
halogenated c o m p o u n d s , ethers, amines, a n d water, the n i n e
basic properties of heavy h y d r o c a r b o n s from all h y d r o c a r b o n
parameters i n Eq. (2.47) were d e t e r m i n e d a n d are given in
groups in the C6-C50 range [65].
Table 2.10. I n using Eq. (2.47), the c o n s t a n t d should not be
Tc, Pc, Vc, I, d 2 0 - - a l [exp(blTb + c l S G + d l T b S G ) ] T~1 SG fi m i s t a k e n with p a r a m e t e r d20 used for liquid density at 20~
(2.46a) As in the other equations i n this chapter, values of Tb a n d Tc
are in kelvin, Pc is in bar, a n d Vc is in cma/g. P a r a m e t e r d20 is
Tb, Tc, Pc, Vc, I, d20 =a2[exp(bEM+CESG the liquid density at 20~ a n d 1.0133 b a r in g/cm 3. For light
(2.46b) +d2MSG)] M e2 SG f2 gases such as m e t h a n e (C1) or ethane (C2) in which they are in
the gaseous state at the reference conditions, a fictitious value
where Vc in these relations is in cm3/mol. Constants a l - f l of dE0 was o b t a i n e d t h r o u g h the extrapolation of density val-
a n d a2-f2 in these relations are given i n Table 2.9. These cor- ues at lower t e m p e r a t u r e given by Reid et al. [4]. The values
relations are r e c o m m e n d e d for h y d r o c a r b o n s a n d p e t r o l e u m of d20 for some gases f o u n d i n this m a n n e r are as follows:
fractions in the c a r b o n n u m b e r range of C20-C50. Although a m m o n i a , NH3 (0.61); n i t r o u s oxide, N20 (0.79); methane,
these equations m a y be used to predict physical properties C1 (0.18); ethane, C2 (0.343); propane, C3 (0.5); n-butane, nC4
of h y d r o c a r b o n s in the range of C6-C20, if the system does (0.579); isobutane, iC4 (0.557); nitrogen, N2 (0.135); oxygen,
n o t c o n t a i n heavy h y d r o c a r b o n s Eqs. (2.38) a n d (2.40) are 02 (0.22); hydrogen sulfide, H2S (0.829); a n d hydrogen chlo-
recommended. ride, HC1 (0.837). I n some references different values for liq-
uid densities of some of these c o m p o u n d s have been reported.
For example, a value of 0.809 g/cm 3 is reported as the den-
2.3.4 Extension of Proposed Correlations
sity of N2 at 15.5~ a n d 1 a t m by several authors in reservoir
to Nonhydrocarbon Systems
engineering [48, 51]. This value is very close to the density of
E q u a t i o n s (2.38) a n d (2.40) c a n n o t be applied to systems con- N2 at 78 K [4]. The critical t e m p e r a t u r e of N2 is 126.1 K a n d
t a i n i n g hydrocarbons, such as m e t h a n e a n d ethane, or hydro-
gen sulfide. These equations are useful for h y d r o c a r b o n s with
c a r b o n n u m b e r s above Cs a n d are not applicable to n a t u r a l TABLE 2.10--Constants for Eq. (2.47).
o~ Vc
gases or refinery gases. E s t i m a t i o n of the properties of n o n h y - Constants Tc. K Pc, MPa cm3'/g
d r o c a r b o n systems is b e y o n d the objective of this book. But a 1.60193 10.74145 -8.84800
in reservoir fluids, c o m p o u n d s such as light h y d r o c a r b o n s b 0.00558 0.07434 -0.03632
or H2S a n d CO2 m a y be present. To develop a generalized c -0.00112 -0.00047 -0.00547
correlation in the form of Eq. (2.40) that includes n o n h y d r o - d -0.52398 -2.10482 0.16629
e 0.00104 0.00508 -0.00028
carbons, usually a third p a r a m e t e r is needed to consider the f -0.06403 -1.18869 0.04660
effects of polarity. I n fact Vetere [66] has defined a polarity g 0.93857 -0.66773 2.00241
factor in terms of the molecular weight a n d boiling p o i n t to h -0.00085 -0.01154 0.00587
predict properties of polar c o m p o u n d s . E q u a t i o n (2.40) was i 0.28290 1.53161 -0.96608
 2. C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P U R E H Y D R O C A R B O N S
therefore at temperature 288 or 293 K it cannot be a liquid
and values reported for density at these temperatures are fic-
titious. In any case the values given here for density of N2,
CO2, C1, Ca, and H2S should not be taken as real values and
they are only recommended for use in Eq. (2.47). It should
be noted that dE0 is the same as the specific gravity at 20~
in the SI system (d4a~ This equation was developed based on
the fact that nonhydrocarbons are mainly polar compounds
and a two-parameter potential energy relation cannot rep-
resent the intermolecular forces between molecules, there-
fore a third parameter is needed to characterize the system.
This method would be particularly useful to estimate the bulk
properties of petroleum fluids containing light hydrocarbons
as well as nonhydrocarbon gases. Evaluation of this method
is presented in Section 2.9.
2.4 PREDICTION OF MOLECULAR WEIGHT,
BOILING POINT, AND SPECIFIC GRAVITY
Molecular weight, M, boiling point, Tb, and specific gravity,
SG, are perhaps the most important characterization param-
eters for petroleum fractions and many physical properties
may be calculated from these parameters. Various methods
commonly used to calculate these properties are presented
here. As mentioned before, the main application of these cor-
relations is for petroleum fractions when experimental data
are not available. For pure hydrocarbons either experimental
data are available or group contribution methods are used to
estimate these parameters [4]. However, methods suggested
in Chapter 3 to estimate properties of petroleum fractions are
based on the method developed from the properties of pure
hydrocarbons in this chapter.
2.4.1 Prediction of Molecular Weight
For pure hydrocarbons from homologous groups, Eq. (2.42)
can be reversed to obtain the molecular weight from other
properties. For example, if Tb is available, M can be estimated
from the following equation:
(2.48) M= [a - ln(Tb~ -- Tb)] /
where values of a, b, c, and Tboo are the same constants as
those given in Table 2.6 for the boiling point. For example,
for n-alkanes, M can be estimated as follows:
(2.49) Mp - ~ / ~ l [6 98291 - ln(1070 - Tb)]} 3/2
[0.02013 '
in which Mp is molecular weight of n-alkane (n-paraffins)
whose normal boiling point is Tb. Values obtained from
Eq. (2.49) are very close to molecular weight of n-alkanes.
Similar equations can be obtained for other hydrocarbon
groups by use of values given in Table 2.6. Once M is deter-
mined from Tb, then it can be used with Eq. (2.42) to obtain
other properties such as specific gravity and critical constants.
2.4.1.1 R i a z i - D a u b e r t M e t h o d s
The methods developed in the previous section are commonly
used to calculate molecular weight from boiling point and
55
specific gravity. Equation (2.38) for molecular weight is [28]
(2.50) M = 1.6607 x 10-4T21962SG -1"0164
This equation fails to properly predict properties for hydro-
carbons above C2s. This equation was extensively evaluated
for various coal liquid samples along with other correlations
by Tsonopoulos et al. [34]. They recommended this equation
for the estimation of the molecular weight of coal liquid frac-
tions. Constants in Eq. (2.40) for molecular weight, as given
in Table 2.5, were modified to include heavy hydrocarbons up
to molecular weight of 700. The equation in terms of Tb and
SG becomes
M = 42.965[exp(2.097 x 10-4Tb -- 7.78712SG
(2.51) + 2.08476 x 10-3TbSG)]Tbl26~176 4"98308
This equation can be applied to hydrocarbons with molecular
weight ranging from 70 to 700, which is nearly equivalent to
boiling point range of 300-850 K (90-1050~ and the API
gravity range of 14.4-93. These equations can be easily con-
verted in terms of Watson K factor (Kw) and API degrees
using their definitions through Eqs. (2.13) and (2.4). A graph-
ical presentation of Eq. (2.51) is shown in Fig. 2.4. (Equation
(2.51) has been recommended by the API as it will be dis-
cussed later.) Equation (2.51) is more accurate for light frac-
tions (M < 300) with an %AAD of about 3.5, but for heavier
fractions the %AAD is about 4.7. This equation is included in
the API-TDB [2] and is recognized as the standard method
of estimating molecular weight of petroleum fractions in the
industry.
For heavy petroleum fractions boiling point may not be
available. For this reason Riazi and Daubert [67] developed
a three-parameter correlation in terms of kinematic viscosity
based on the molecular weight of heavy fractions in the range
of 200-800:
[. (-1.2435+1.1228SG) .(3.4758-3.038SG)'1 SG-0.6665
M = 223.56 l_v38(loo) u99(21o) j
(2.52)
The three input parameters are kinematic viscosities (in cSt)
at 38 and 98.9~ (100 and 210~ shown by v38000) and 1)99(210),
respectively, and the specific gravity, SG, at 15.5~ It should
be noted that viscosities at two different temperatures repre-
sent two independent parameters: one the value of viscosity
and the other the effect of temperature on viscosity, which is
another characteristic of a compound as discussed in Chap-
ter 3. The use of a third parameter is needed to character-
ize complexity of heavy hydrocarbons that follow a three-
parameter potential energy relation. Equation (2.52) is only
recommended when the boiling point is not available. In
a case where specific gravity is not available, a method is
proposed in Section 2.4.3 to estimate it from viscosity data.
Graphical presentation of Eq. (2.52) is shown in Fig. 2.5 in
terms of API gravity. To use this figure, based on the value of
v3m00) a point is determined on the vertical line, then from
v a l u e s of I)99(210) and SG, another point on the chart is speci-
fied. A line that connects these two points intersects with the
line of molecular weight where it may be read as the estimated
value.
56 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

800
o n-alkane
700

600

500

400

300

200

100 y _.
~ _ _
200~
150~
30oC 100~
0 i ! t , I } i

0.60 0.70 0.80 0.90 1.00 1.10

Specific Gravity

FIG. 2.4---Estimation of molecular weight from Eq. (2.51).

2.4.1.2 ASTM Method Solution--In using Eq. (2.52) three p a r a m e t e r s of v38o00),

ASTM D 2502 m e t h o d [68] provides a c h a r t to calculate the
m o l e c u l a r weight of viscous oils using the k i n e m a t i c viscosi-
ties m e a s u r e d at 100~ (38~ a n d 210~ (99~ The m e t h o d
was e m p i r i c a l l y developed by H i r s c h l e r in 1946 [69] a n d is
p r e s e n t e d b y the following equation.
(2.53) M = 180 + K(Ha80oo) + 60)
where
K = 4.145 - 1.733 l o g t o ( V S F - 145)
VSF = / - / 3 8 ( 1 0 0 ) - H 9 9 ( 2 1 0 )
H = 870 lOglo[lOglo(v + 0.6)] + 154
in w h i c h v is the k i n e m a t i c viscosity in cSt. This e q u a t i o n
was developed s o m e 60 years ago a n d requires k i n e m a t i c
viscosities at 38 a n d 99~ in cSt as the only i n p u t p a r a m -
eters. The H i r s c h l e r m e t h o d was i n c l u d e d in the API-TDB
in 1964 [2], b u t in the 1987 revision of API-TDB it was re-
p l a c e d b y Eq. (2.52). Riazi a n d D a u b e r t [67] extensively com-
p a r e d Eq. (2.52) with the H i r s c h l e r m e t h o d a n d they f o u n d
t h a t for s o m e 160 fractions in the m o l e c u l a r weight range of
200-800 the p e r c e n t average absolute deviation (%AAD) for
these m e t h o d s were 2.7% a n d 6.9%, respectively. Even if the
constants of the H i r s c h l e r c o r r e l a t i o n were r e o b t a i n e d f r o m
the d a t a b a n k u s e d for the evaluations, the a c c u r a c y of the
m e t h o d i m p r o v e d only slightly from 6.9 to 6.1% [67].
Example 2 . 4 - - T h e viscosity a n d o t h e r p r o p e r t i e s of 5-n-
butyldocosane, C26H54, as given in API RP-42 [18] are M =
366.7, SG -- 0.8099, v38000) = 11.44, a n d 1)99(210) = 3.02 cSt.
Calculate the m o l e c u l a r weight with %AD f r o m the API
m e t h o d , Eq. (2.52), a n d the H i r s c h l e r m e t h o d (ASTM 2502),
Eq. (2.53).
P99(210), a n d SG are needed.
M = 223.56 x [11.44 (-1"2435+1"1228x0"8099)• 3.02 (3"475a-a'038x0"8099)]
• % A D = 4 . 5 % . F r o m Eq. (2.53):
H38 = 183.3, H 9 9 ~--- - 6 5 . 8 7 , VSF---249.17, K = 0.6483, M =
337.7, %AD = 7.9%. #
2.4.1.3 API Methods
The API-TDB [2] a d o p t e d m e t h o d s developed b y Riazi a n d
D a u b e r t for the e s t i m a t i o n of the m o l e c u l a r weight of hydro-
c a r b o n systems. I n the 1982 edition of API-TDB, a modified
version of Eq. (2.38) was included, b u t in its latest editions
(from 1987 to 1997) Eqs. (2.51) a n d (2.52) are i n c l u d e d after
r e c o m m e n d a t i o n s m a d e b y the API-TDB Committee.
2.4.1.4 Lee--Kesler Method
The m o l e c u l a r weight is related to boiling p o i n t a n d specific
gravity t h r o u g h a n e m p i r i c a l r e l a t i o n as follows [13]:
M = - 1 2 2 7 2 . 6 + 9486.4SG + (8.3741 - 5.9917SG )Tb
+ (1 - 0.77084SG - 0.02058SG 2)
(2.54) x (0.7465 - 222.466/Tb)lO7/Tb
+ (1 -- 0.80882SG + 0.02226SG 2)
x (0.3228 - 17.335/Tb)1012/T~
High-molecular-weight d a t a were also u s e d in o b t a i n i n g the
constants. The correlation is r e c o m m e n d e d for use u p to a
boiling p o i n t o f a b o u t 750 K (~850~ Its evaluation is s h o w n
in Section 2.9.
 2. CHARACTERIZATION AND P R O P E R TIE S OF PURE HYDROCARBONS 57

_10000
"~ooo API G r a v i t y
-8000
_~7ooo
-6000 20
=5000 4 BOO
~'4000
~ 3000

--- 2 0 0 0
700
m
-'1000
900
800
7OO
600
500
600
400
o
3OO

oT - -
-200

._~ 5 0 0
~0 (D
100
90
o
o
.(~
>
%
80
70 o 0
60 o
50 E
c 4OO
40 5

30 10

-20

300

!0
g
8
7
6
5
200
4
3
_=

100

E ,

FIG. 2.5---Estimation of molecular weight from Eq. (2.52). Taken from Ref. [67] with per-
mission.

2.4.1.5 Goossens Correlation Eq. (2.38) b u t with a variable b a n d c = - 1 . P a r a m e t e r b is

Most recently Goossens [61] c o r r e l a t e d M to Tb a n d d20 in the
following f o r m using the d a t a on 40 p u r e h y d r o c a r b o n s a n d
23 p e t r o l e u m fractions:
(2.55) M = 0.01077Tb~/d] ~
w h e r e /3 = 1.52869 + 0.06486 ]n[Tb/(1078 -- Tb)]. I n s p e c t i o n
of this e q u a t i o n shows t h a t it has the s a m e structure as
c o n s i d e r e d as a function of Tb, while SG in Eq. (2.38) is re-
p l a c e d b y d~~ the specific gravity at 20/4~ (d~~ is the s a m e
as d20 in g/cm3). The d a t a b a n k u s e d to develop this equa-
tion covers the c a r b o n r a n g e of C5-C120 (M ~ 70-1700, Tb
300-1000 K, a n d d ~ 0.63-1.08). F o r the s a m e 63 d a t a p o i n t s
u s e d to o b t a i n the constants of Eq. (2.55), the average e r r o r
was 2.1%. However, p r a c t i c a l a p p l i c a t i o n of Eq. (2.55) is lim-
ited to m u c h lower m o l e c u l a r weight fractions b e c a u s e heavy
58 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
fraction distillation data is not usually available. When d20 is
not available it may be estimated from SG using the method
given in Section 2.6.1.
2.4.1.6 Other Methods
Twu [30] proposed a set of correlations for the calculation of
M, Tc, Pc, and Vc of hydrocarbons. Because these correlations
are interrelated, they are all given in Section 2.5.1. The com-
puterized Winn method is given by Eq. (2.93) in Section 2.5.1
and in the form of chart in Fig. 2.12.
Example 2.5--For n-Butylbenzene estimate the molecular
weight from Eqs. (2.50), (2.51), (2.54), and (2.55) using the
input data from Table 2.1
Solution--From Table 2.1, n-butylbenzene has Tb = 183.3~
SG-- 0.8660, d = 0.8610, and M = 134.2. Applying various
equations we obtain the following: from Eq. (2.50), M = 133.2
with AD-- 0.8%, Eq. (2.51) gives M = 139.2 with AD = 3.7%,
Eq. (2.54) gives M = 143.4 with AD-- 6.9%, and Eq. (2.55)
gives M = 128.7 with AD = 4.1%. For this pure and light hy-
drocarbon, Eq. (2.50) gives the lowest error because it was
mainly developed from the molecular weight of pure hydro-
carbons while the other equations cover wider range of molec-
ular weight because data from petroleum fractions were also
used in their development, t Specific gravity of hydrocarbons and petroleum fractions is
2.4.2 Prediction of Normal Boiling Point
2.4.2.1 Riazi-Daubert Correlations
These correlations are developed in Section 2.3. The best in-
put pair of parameters to predict boiling point are (M, SG) or
(M, 1). For light hydrocarbons and petroleum fractions with
molecular weight in the range of 70-300, Eq. (2.40) may be
used for boiling point:
Tb = 3.76587[exp(3.7741 x 10-3M + 2.98404SG
(2.56) -4.25288 • 10-3MSG)]M~176 -1'58262
For hydrocarbons or petroleum fractions with molecular
weight in the range of 300-700, Eq. (2.46b) is recommended:
Tb = 9.3369[exp(1.6514 x 10-4M + 1.4103SG
(2.57) - 7.5152 x 10-4MSG)]M~ -0"7276
Equation (2.57) is also applicable to hydrocarbons having
molecular weight range of 70-300, with less accuracy. Esti-
mation of the boiling point from the molecular weight and
refractive index parameter ( I ) is given by Eq. (2.40) with con-
stants in Table 2.5. The boiling point may also be calculated
through Kw and API gravity by using definitions of these pa-
rameters given in Eqs. (2.13) and (2.4).
2.4.2.2 Soreide Correlation
Based on extension of Eq. (2.56) and data on the boiling point
of some C7+ fractions, Soreide [51, 52] developed the follow-
ing correlation for the normal boiling point of fractions in the
range of 90-560~
Tb = 1071.28 -- 9.417 x 104 exp(--4.922 x 10-aM
(2.58) - 4.7685SG + 3.462 x 10-3MSG) M-~176 TM
This relation is based on the assumption that the boiling point
of extremely large molecules (M --~ c~) approaches a finite
value of 1071.28 K. Soreide [52] compared four methods for
the prediction of the boiling point of petroleum fractions:
(1) Eq. (2.56), (2) Eq. (2.58), (3) Eq. (2.50), and (4) Twu
method given by Eqs. (2.89)-(2.92). For his data bank on
boiling point of petroleum fractions in the molecular weight
range of 70-450, he found that Eq. (2.50) and the Twn cor-
relations overestimate the boiling point while Eqs. (2.56)
and (2.58) are almost identical with AAD of about 1%. Since
Eq. (2.56) was originally based on hydrocarbons with a molec-
ular weight range of 70-300, its application to heavier com-
pounds should be taken with care. In addition, the database
for evaluations by Soreide was the same as the data used to
derive constants in his correlation, Eq. (2.58). For heavier hy-
drocarbons (M > 300) Eq. (2.57) may be used.
For pure hydrocarbons from different homologous families
Eq. (2.42) should be used with constants given in Table 2.6 for
Tb to estimate boiling point from molecular weight. A graph-
ical comparison of Eqs. (2.42), (2.56), (2.57), and (2.58) for
n-alkanes from C5 to C36 with data from API-TDB [2] is shown
in Fig. 2.6.
2.4.3 Prediction of Specific Gravity/API Gravity
normally available because it is easily measurable. Specific
gravity and the API gravity are related to each other through
Eq. (2.4). Therefore, when one of these parameters is known
the other one can be calculated from the definition of the API
gravity. Several correlations are presented in this section for
the estimation of specific gravity using boiling point, molec-
ular weight, or kinematic viscosity as the input parameters.
2.4.3.1 Riazi-Daubert Methods
These correlations for the estimation of specific gravity re-
quire Tb and I or viscosity and CH weight ratio as the input
parameters (Eq. 2.40). For light hydrocarbons, Eq. (2.40) and
Table 2.5 can be used to estimate SG from different input
parameters such as Tb and I.
SG = 2.4381 x 107 exp(-4.194 x 10-4Tb -- 23.55351
(2.59) + 3.9874 x lO-3Tbl)Tb~
where Tb is in kelvin. For heavy hydrocarbons with molecular
weight in the range 300-700, the following equation in terms
of M and I can be used [65]:
SG = 3.3131 x 104 exp(-8.77 x 1 0 - a M - 15.0496I
(2.60) + 3.247 x lO-3MI)M-~176 4"9557
Usually for heavy fractions, Tb is not available and for this rea-
son, M and I are used as the input parameters. This equation
also may be used for hydrocarbons below molecular weight
of 300, if necessary. The accuracy of this equation is about 0.4
%AAD for 130 hydrocarbons in the carbon n u m b e r range of
C7-C50 (M ~ 70-700).
For heavier fractions (molecular weight from 200 to 800)
and especially when the boiling point is not available the fol-
lowing relation in terms of kinematic viscosities developed by
 2. CHARACTERIZATION AND P R O P E R TIE S OF PURE H YD R O CA R B O N S 59

600
o

o API Data 9 r
. . . . . . . Eq. 2.42 ~ ,,~,~
500 .... Eq. 2.56

400
P
300

200

I00

0 1 i I ii i I i I I L J I I I I I I I i I I I i

0 5 10 15 20 25 30 35 4O
Carbon Number

FIG. 2.6--Estimation of boiling point of n-alkanes from various methods.

50 0.78

40

0,88
30

o
o
20
0.98

l0

0 1.08
10 100 1000
Kinematic Viscosity at 37.8 ~ cSt
FIG. 2.7--API gravity and viscosity of heavy hydrocarbon fractions by Eq. (2.61).
60 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

Riazi and Daubert may be used [67]: Riazi-Daubert or Riazi methods. These equations are recom-
(2.61) SG ---~ 0.1157 r -0.16161
mended only for hydrocarbons in the molecular weight range
0.7717 [v38000) ] • [1299(210
)] of 70-300 and have been widely used in industry [2, 47, 49, 51,
in which v38~100~and ~d99(210) are kinematic viscosities in cSt at 54, 70]. However, these correlations were replaced with more
i00 and 210~ (37.8 and 98.9~ respectively. Equation (2.61) accurate correlations presented by Eq. (2.40) and Table 2.5 in
is shown in Fig. 2.7 and has also been adopted by the API and terms of Tb and SG as given below:
is included in 1987 version of API-TDB [2]. This equation
Tc = 9.5233[exp(-9.314 x 10-4Tb - - 0 . 5 4 4 4 4 2 S G
gives an AAD of about 1.5% for 158 fractions in the molecular
weight range of 200-500 (~SG range of 0.8-1.1). (2.65) +6.4791 • 1 0 - 4 T b S G ) ] T ~ 1 7 6 0"53691
For coal liquids and heavy residues that are highly aro-
matic, Tsonopoulos et al. [58] suggest the following relation Pc = 3.1958 x 105[exp(-8.505 x 10-3Tb - - 4 . 8 0 1 4 S G
in terms of normal boiling point (Tb) for the estimation of (2.66) + 5.749 x 10-3TbSG)]Tb~ 4"0846
specific gravity.
These correlations were also adopted by the API and have
SG = 0.553461 + 1.15156To - 0.708142To 2 + 0.196237T3 been used in m a n y industrial computer softwares under the
(2.62) API method. The same limitations and units as those for Eqs.
(2.63) and (2.64) apply to these equations. For heavy hydro-
where To = (1.8Tb- 459.67) in which Tb is in kelvin. This
carbons (>C20) the following equations are obtained from
equation is not recommended for pure hydrocarbons or
Eq. (2.46a) and constants in Table 2.9:
petroleum fractions and has an average relative deviation of
about 2.5% for coal liquid fractions [58]. For pure homolo- Tc = 35.9413[exp(-6.9 x 10-4Tb -- 1.4442SG
gous hydrocarbon groups, Eq. (2.42) with constants given in (2.67) +4.91 X 10-4TbSG)]Tb~ 1"2771
Table 2.6 for SG can be used. Another approach to estimate
specific gravity is to use the Rackett equation and a known
Pc = 6.9575[exp(-1.35 • 10-2Zb -- 0.3129SG
density data point at any temperature as discussed in Chap-
ter 5 (Section 5.8). A very simple and practical method of (2.68) + 9.174 x 1 0 - 3 T b S G ) ] T ~ -0"6807
estimating SG from density at 20~ d, is given by Eq. (2.110),
If necessary these equations can also be used for hydrocar-
which will be discussed in Section 2.6.1. Once SG is esti-
bons in the range of C5-C20 with good accuracy. Equation
mated the API gravity can be calculated from its definition,
(2.67) predicts values of Tc from C5 to Cs0 with %AAD of 0.4%,
i.e., Eq. (2.4).
but Eq. (2.68) predicts Pc with AAD of 5.8%. The reason for
this high average error is low values of Pc (i.e, a few bars) at
higher carbon numbers which even a small absolute deviation
2.5 P R E D I C T I O N OF CRITICAL shows a large value in terms of relative deviation.
P R O P E R T I E S A N D A C E N T R I C FACTOR
2.5.1.2 API Methods
Critical properties, especially the critical temperature and
The API-TDB [2] adopted methods developed by Riazi and
pressure, and the acentric factor are important input param-
Daubert for the estimation of pseudocritical properties of
eters for EOS and generalized correlations to estimate phys-
petroleum fractions. In the 1982 edition of API-TDB, Eqs.
ical and thermodynamic properties of fluids. As shown in
(2.63) and (2.64) were recommended for critical temperature
Chapter 1 even small errors in prediction of these proper-
and pressure of petroleum fractions, respectively, but in its
ties greatly affect calculated physical properties. Some of the
editions from 1987 to 1997, Eqs. (2.65) and (2.66) are included
methods widely used in the petroleum industry are given in
after evaluations by the API-TDB Committee. For pure hy-
this section. These procedures, as mentioned in the previous
drocarbons, the methods recommended by API are based on
sections, are mainly developed based on critical properties of
group contribution methods such as Ambrose, which requires
pure hydrocarbons in which validated experimental data are
the structure of the compound to be known. These methods
available only up to C18. The following correlations are given
are of minor practical use in this book since properties of
in terms of boiling point and specific gravity. For other in-
pure compounds of interest are given in Section 2.3 and for
put parameters, appropriate correlations given in Section 2.3
petroleum fractions the bulk properties are used rather than
should be used.
the chemical structure of individual compounds.

2.5.1 Prediction of Critical Temperature 2.5.1.3 Lee-Kesler Method

and Pressure Kesler and Lee [12] proposed correlations for estimation of
Tc and Pc similar to their correlation for molecular weight.
2.5.1.1 Riazi-Daubert Methods
Simplified equations to calculate T~ and Pc of hydrocarbons Tc = 189.8 + 450.6 SG + (0.4244 + 0.1174 SG)Tb
in the range of C5-C20 are given by Eq. (2.38) as follows [28]. (2.69) +(0.1441 - 1.0069 SG)105/Tb
(2.63) Tc = 19.06232Tb~ ~
In Pc = 5.689 - 0.0566/SG
(2.64) Pc = 5.53027 • 107Tb2"3125SG2"3201 - (0.43639 + 4.1216/SG + 0.21343/SG 2) x 10-3Tb
where Tc and Tb are in kelvin and Pc is in bar. In the litera- (2.70) + (0.47579 + 1.182/SG + 0.15302/SG 2) x 10 -6 • T~
ture, Eqs. (2.50), (2.63), and (2.64) are usually referred to as - (2.4505 + 9.9099/SG 2) • 10 -1~ x T~
 2. CHARACTERIZATION AND P R O P E R T I E S OF P U R E H Y D R O C A R B O N S 61

w h e r e Tb a n d Tc are in kelvin a n d Pc is in bar. I n these equa- i n c l u d e d in s o m e references [49Z1. However, the Twu correla-
tions a t t e m p t s were m a d e to keep internal consistency a m o n g tions a l t h o u g h b a s e d on the s a m e f o r m a t as the KLS o r LC
Tc a n d Pc that at Pc equal to 1 atm, Tc is c o i n c i d e d with nor- require i n p u t p a r a m e t e r s of Tb a n d SG a n d are a p p l i c a b l e to
m a l boiling point, Tb. The correlations were r e c o m m e n d e d h y d r o c a r b o n s b e y o n d C20. F o r heavy h y d r o c a r b o n s s i m i l a r
by the a u t h o r s for the m o l e c u l a r range of 70-700 (~C5-C50). to the a p p r o a c h of L e e - K e s l e r [12], Twu [30] used the crit-
However, the values of Tc a n d Pc for c o m p o u n d s with c a r b o n ical p r o p e r t i e s b a c k calculated f r o m v a p o r p r e s s u r e d a t a to
n u m b e r s greater t h a n Cls used to develop the above correla- e x p a n d his d a t a b a n k on the critical c o n s t a n t s of p u r e hydro-
tions were not b a s e d on e x p e r i m e n t a l evidence. c a r b o n c o m p o u n d s . F o r this r e a s o n the Twu correlations have
f o u n d a w i d e r range of application. The Twu correlations for
2.5.1.4 Cavett Method the critical properties, specific gravity, and m o l e c u l a r weight
Cavett [26] developed e m p i r i c a l correlations for Tc a n d Pc in of n-alkanes are as follows:
t e r m s of boiling p o i n t a n d API gravity, w h i c h are still available
T~~ = Tb(0.533272 + 0.34383 x 10 -3 • Tb
in s o m e process s i m u l a t o r s as an o p t i o n a n d in s o m e cases
give g o o d estimates of Tc a n d Pc for light to m i d d l e distillate +2.52617 x 10 -7 • T~ - 1.658481 • 10 -1~ x T3
p e t r o l e u m fractions. (2.73) +4.60773 x 1024 x Tb-13)-1

Tc = 426.7062278 + (9.5187183 x 10-~)(1.8Tb -- 459.67)

(2.74) ot = 1 - Tt,/T~
- (6.01889 x 10-4)(1.8Tb -- 459.67) 2
-- (4.95625 x 10-3)(API)(1.8Tb - 459.67) P~ = (1.00661 + 0.31412ot 1/2 + 9.161063
(2.71) +(2.160588 x 10-7)(1.8Tb - 459.67) 3 (2.75) + 9.504132 + 27.358860t4) 2
+ (2.949718 • 10-6)(API)(1.8Tb - 459.67) 2
+(1.817311 x 10-8)(APIZ)(1.8Tb- 459.67) 2 V~~ = (0.34602 + 0.301710~ + 0.93307ot 3 + 5655.414314) -s
(2.76)
log(Pc) = 1.6675956 + (9.412011 x 10-4)(l.8Tb - 459.67)
- (3.047475 • 10-6)(1.8Tb -- 459.67) 2 SG ~ = 0.843593 - 0.128624ot - 3.361590t 3 - 13749.5312
-(2.087611 x 10-5)(API)(1.8Tb- 459.67) (2.77)
(2.72) + (1.5184103 x 10-9)(1.8Tb -- 459.67) 3
+ (1.1047899 x 10-8)(API)(1.8Tb - 459.67) 2 Tb = exp(5.12640 + 2.71579fl -- 0.286590fl 2 -- 39.8544/fl
- (4.8271599 x 10-8)(API2)(1.8Tb - 459.67) (2.78) --0.122488/fl 2) -- 13.7512fl + 19.6197fl 2
+ (1.3949619 x 10-1~ - 459.67) 2 w h e r e Tb is the boiling p o i n t of h y d r o c a r b o n s in kelvin a n d
3 = l n ( M ~ in w h i c h M ~ is the m o l e c u l a r weight n-alkane ref-
I n these relations Pc is in b a r while Tc a n d Tb are in kelvin a n d
erence c o m p o u n d . Critical pressure is in b a r a n d critical vol-
the API gravity is defined in t e r m s of specific gravity t h r o u g h
u m e is in cm3/mol. Data on the p r o p e r t i e s of n-alkanes from
Eq. (2.4). Terms (1.8Tb - 459.67) c o m e from the fact t h a t the
C1 to C100 were u s e d to o b t a i n the constants in the a b o v e rela-
unit of Tb in the original relations was in degrees fahrenheit.
tions. F o r heavy h y d r o c a r b o n s b e y o n d C20, the values of t h e
critical p r o p e r t i e s o b t a i n e d f r o m v a p o r p r e s s u r e d a t a were
2.5.1.5 Twu Method for Tc, Pc, Vc, and M
used to o b t a i n the constants. The a u t h o r o f these correla-
Twu [30] initially c o r r e l a t e d critical p r o p e r t i e s (To, Pc, Vc), tions also indicates that there is i m e r n a l consistency b e t w e e n
specific gravity (SG), and m o l e c u l a r weight (M) of n-alkanes Tc a n d Pc as the critical t e m p e r a t u r e a p p r o a c h e s the boiling
to the boiling p o i n t (Tb). Then the difference b e t w e e n spe- point. E q u a t i o n (2.78) is implicit irt calculating M ~ f r o m Tb. To
cific gravity of a h y d r o c a r b o n from o t h e r g r o u p s (SG) a n d solve this equation by iteration a starting value c a n be f o u n d
specific gravity of n-alkane (SG ~ was used as the s e c o n d pa- from the following relation:
r a m e t e r to correlate p r o p e r t i e s of h y d r o c a r b o n s f r o m differ-
ent groups. This type of correlation, k n o w n as a p e r t u r b a t i o n (2.79) M ~ = Tb/(5.8 -- 13'.0052Tb)
expansion, was first i n t r o d u c e d by K e s l e r - L e e - S a n d l e r (KLS)
[71] a n d later u s e d by Lin a n d Chao [72] to correlate critical F o r o t h e r h y d r o c a r b o n s a n d petro][eum fractions the r e l a t i o n
p r o p e r t i e s of h y d r o c a r b o n s using n-alkane as a reference fluid for the e s t i m a t i o n of To, Pc, Vc, and M are as follows:
a n d the specific gravity difference as the correlating p a r a m - Critical temperature
eter. However, KLS correlations d i d not find practical appli-
(2.80) Tc = Tc[(1 + 2fr)/(1 - 2fr)] 2
cation b e c a u s e they defined a n e w t h i r d p a r a m e t e r s i m i l a r to
the acentric factor w h i c h is not available for p e t r o l e u m mix-
tures. Lin a n d Chao (LC) c o r r e l a t e d Tc, ln(Pc), w, SG, a n d Tb fr = A S G T [ - 0.27016/T~/2
of n-alkanes from CI to C20 to m o l e c u l a r weight, M. These (2.81) + (0.0398285 - 0.706691/T~/2)ASGr]
p r o p e r t i e s for all o t h e r h y d r o c a r b o n s in the s a m e m o l e c u l a r
weight were c o r r e l a t e d to the difference in Tb a n d SG of the
(2.82) A S G r = exp[5(SG ~ - SG)] - 1
s u b s t a n c e of interest with that of n-alkane. Therefore, LC cor-
relations r e q u i r e three i n p u t p a r a m e t e r s of Tb, SG, a n d M for Critical volume
each property. E a c h correlation for each p r o p e r t y c o n t a i n e d
as m a n y as 33 n u m e r i c a l constants. These correlations are (2.83) Vc = vf[(1 + 2 fv)/(1 - 2 f v ) ] 2
62 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
fv = ASGv[O.347776/T~/2
(2.84) + ( - 0.182421 + 2.248896/T~/2)ASGv]
(2.85) ASGv = exp[4(SG ~ - SG2)] - 1
Critical pressure
(2.86) Pc = P~(TclT~) x (Vc/Vc)[(1 + 2fv)l(1 -2fp)] 2
fe = ASGe[(2.53262 - 34.4321/T~/2 - 2.30193Tb/1000)
+( -- 11.4277 + 187.934/Tb1/2 + 4.11963Tb/lOOO)ASGv]
(2.87)
(2.88) ASGv = exp[0.5(SG ~ - SG)] - 1
Molecular weight
(2.89) ln(M) = (ln M~ + 2 f~)/(1 - 2fM)] 2
f?a = ASGm[x + (-0.0175691 + 0.143979/T~/2)ASGM]
(2.90)
(2.91) X --- ]0.012342 - 0.244541/T1/2[
(2.92) ASGM = exp[5(SG ~ - SG)] - 1
In the above relations Tb and Tc are in kelvin, Vc is in cm3/mol,
and Pc is in bar. One can see that these correlations should
be solved simultaneously because they are highly interrelated
to each other and for this reason relations for estimation of
M and Vcbased on this method are also presented in this part.
Example 2.6---Estimate the molecular weight of n-eicosane
(C20H42) from its normal boiling point using Eq. (2.49) and
the Twu correlations.
Solution--n-Eicosane is a normal paraffin whose molecular
weight and boiling point are given in Table 2. I as M = 282.55
and Tb = 616.93 K. Substituting Tb in Eq. (2.49) gives M =
282.59 (%AD = 0.01%). Using the Twu method, first an initial
guess is calculated through Eq. (2.79) as M ~ = 238 and from
iteration the final value of M ~ calculated from Eq. (2.78) is
281.2 (%AD = 0.48%). Twu method for estimation of proper-
ties of hydrocarbons from other groups is shown later in the
next example. r range of C5-C20 is given by Eq. (2.38) as follows.
2.5.1.6 Winn-Mobil Method
Winn [25] developed a convenient nomograph to estimate var-
ious physical properties including molecular weight and the
pseudocritical pressure for petroleum fractions. Mobil [73]
proposed a similar nomograph for the estimation of pseudo-
critical temperature. The input data in both nomographs are
boiling point (or Kw) and the specific gravity (or API gravity).
As part of the API project to computerize the graphical meth-
ods for estimation of physical properties, these nomographs
were reduced to equation forms for computer applications by
Riazi [36] and were later reported by Sire and Daubert [74].
These empirically developed correlations have forms similar
to Eq. (2.38) and for M, To and Pc are as follows.
(2.93) M = 2.70579 x 10-5T~4966SG-1'174
(2.94) In Tc = -0.58779 + 4.2009T~176 ~176
(2.95) Pc = 6.148341 x 107Tb23177SG2"4853
where Tb and Tc are in kelvin and Pc is in bar. Comparing
values estimated from these correlations with the values from
the original figures gives AAD of 2, 1, and 1.5% for M, To, and
Pc, respectively, as reported in Ref. [36]. In the literature these
equations are usually referred as Winn or Sim-Daubert and
are included in some process simulators. The original Winn
nomograph for molecular weight and some other properties
is given in Section 2.8.
2.5.1.7 Tsonopoulos Correlations
Based on the critical properties of aromatic compounds,
Tsonopoulos et al. [34] proposed the following correlations
for estimation of Tc and Pc for coal liquids and aromatic-rich
fractions.
lOgl0 Tc = 1.20016 + 0.61954(log10 Tb)
(2.96) + 0.48262(1og10 SG) + 0.67365(log10 SG) 2
log10 Pc = 7.37498 - 2.15833(log10 Tb)
(2.97) + 3.35417(log10 SG) + 5.64019(log10 SG) 2
where Tb and Tc are in kelvin and Pc is in bar. These correla-
tions are mainly recommended for coal liquid fractions and
they give average errors of 0.7 and 3.5% for the estimation of
critical temperature and pressure of aromatic hydrocarbons.
2.5.2 Prediction o f Critical Volume
Critical volume, Vo is the third critical property that is not
directly used in EOS calculations, but is indirectly used to
estimate interaction parameters (kii) needed for calculation
of mixture pseudocritical properties or EOS parameters as
will be discussed in Chapter 5. In some corresponding state
correlations developed to estimate transport properties of flu-
ids at elevated pressure, reduced density (Vc/V) is used as the
correlating parameter and values of Vc are required as shown
in Chapter 8. Critical volume is also used to calculate critical
compressibility factor, Zc, as shown by Eq. (2.8).
2.5.2.1 Riazi-Daubert Methods
A simplified equation to calculate Vc of hydrocarbons in the
(2.98) Vc = 1.7842 x 10-4T2"a829SG-1"683
in which Vc is in cma/mol and Tb is in kelvin. When evalu-
ated against more than 100 pure hydrocarbons in the carbon
range of C5--C20 an average error of 2.9% was observed. This
equation may be used up to Css with reasonable accuracy. For
heavier hydrocarbons, Vc is given by Eq. (2.46a) and in terms
of Tb and SG is given as
Vr = 6.2 x 101~ x 10-3Tb -- 28.5524SG
(2.99) + 1.172 x aO-2TbSG)]Tl2~ 17"2074
where Vc is in cm3/mol. Although this equation is recom-
mended for hydrocarbons heavier than C20 it may be used, if
necessary, for the range of C5-C50 in which the AAD is about
2.5%. To calculate Vc from other input parameters, Eqs. (2.40)
and (2.46b) with Tables 2.5 and 2.9 may be used.
 2. CHARACTERIZATION AND PROPERTIES OF PUAE HYDROCARBONS
2.5.2.2 Hall-Yarborough Method
This method for estimation of critical volume follows the gen-
eral form of Eq. (2.39) in terms of M and SG and is given as
[75]:
(2.100) Vc = 1.56 MllSSG -0"7935
Predictive methods in terms of M and SG are usually useful
for heavy fractions where distillation data may not be avail-
able.
2.5.2.3 API Method
In the most recent API-TDB [2], the Reidel method is rec-
ommended to be used for the critical volume of pure hydro-
carbons given in terms of To, Pc, and the acentric factor as
follows:
RTc
(2.101) Vc =
P~[3.72 + 0.26(t~R - 7.00)]
in which R is the gas constant and OR is the Riedel factor given
in terms of acentric factor, o3.
(2.102) Ot R = 5.811 + 4.919o3
In Eq. (2.101), the unit of Vc mainly depends on the units of
To, Pc, and R used as the input parameters. Values of R in
different unit systems are given in Section 1.7.24. To have Vc
in the unit of cma/mol, Tc must be in kelvin and if Pc is in
bar, then the value of R must be 83.14. The API method for
calculation of critical volume of mixtures is based on a mixing
rule and properties of pure compounds, as will be discussed
in Chapter 5. Twu's method for estimation of critical volume
is given in Section 2.5.1.
2.5.3 Prediction of Critical Compressibility Factor
Critical compressibility factor, Zc, is defined by Eq. (2.8) and
is a dimensionless parameter. Values of Zc given in Table 2.1
show that this parameter is a characteristic of each com-
pound, which varies from 0.2 to 0.3 for hydrocarbons in the
range of C1-C20. Generally it decreases with increasing car-
bon number within a homologous hydrocarbon group. Zc is
in fact value of compressibility factor, Z, at the critical point
and therefore it can be estimated from an EOS. As it will be
seen in Chapter 5, two-parameter EOS such as van der Waals
or Peng-Robinson give a single value of Zc for all compounds
and for this reason they are not accurate at the critical re-
gion. Three-parameter EOS or generalized correlations gen-
erally give more accurate values for Zc. On this basis some
researchers correlated Zc to the acentric factor. An example
of such correlations is given by Lee-Kesler [27]:
(2.103) Z~ = 0.2905 - 0.085w
Other references give various versions of Eq. (2.103) with
slight differences in the numerical constants [6]. Another ver-
sion of this equation is given in Chapter 5. However, such
equations are only approximate and no single parameter is
capable of predicting Zc as its nature is different from that of
acentric factor.
Another method to estimate Z~ is to combine Eqs. (2.101)
and (2.102) and using the definition of Zc through Eq. (2.8)
to develop the following relation for Zc in terms of acentric
63
factor, m:
1.1088
(2.104) Zc --
o3+ 3.883
Usually for light hydrocarbons Eq. (2.103) is more accurate
than is Eq. (2.104), while for heavy compounds it is the op-
posite; however, no comprehensive evaluation has been made
on the accuracy of these correlations.
Based on the methods presented in this chapter, the most
appropriate method to estimate Zc is first to estimate To, Pc,
and Vc through methods given in Sections 2.5.1 and 2.5.2 and
then to calculate Zc through its definition given in Eq. (2.8).
However, for consistency in estimating To, Pc, and Vc, one
method should be chosen for calculation of all these three
parameters. Figure 2.8 shows prediction of Zc from various
correlations for n-alkanes from C.5 to C36 and comparing with
data reported by API-TDB [2].
Example 2.7--The critical properties and acentric factor of
n-hexatriacontane (C36H74) are given as follows [20]: Tb =
770.2 K, SG = 0.8172, M = 506.98, Tc = 874.0 K, Pc = 6.8 bar,
Vc = 2090 cm3/mol, Zc = 0.196, and w = 1.52596. Calculate
M, To, Pc, Vc, and Zc from the following methods and for each
property calculate the percentage relative deviation (%D) be-
tween estimated value and other actual value.
a. Riazi-Daubert method: Eq. (2.38)
b. API methods
c. Riazi-Daubert extended method: Eq. (2.46a)
d. Riazi-Sahhaf method for hon~tologous groups, Eq. (2.42),
Pc from Eq. (2.43)
e. Lee-Kesler methods
f. Cavett method (only Tc and Pc), Zc from Eq. (2.104)
g. Twu method
h. Winn method (M, Tc, Pc) and ttall-Yarborough for Vc
i. Tabulate %D for various properties and methods.
Solution--(a) Riazi-Daubert method by Eq. (2.38) for M, To,
Pc, and Vc are given by Eqs. (2.50), (2.63), (2.64), and (2.98).
(b) The API methods for prediction of M, To, Pc, Vc, and
Zc are expressed by Eqs. (2.51), (2.65), (2.66), (2.101), and
(2.104), respectively. (c) The extended Riazi-Daubert method
expressed by Eq. (2.46a) for hydrocarbons heavier than C20
and constants for the critical properties are given in Table 9.
For Tc, Pc, and Vc this method is presented by Eqs. (2.67),
(2.68), and (2.99), respectively. The relation for molecular
weight is the same as the API method, Eq. (2.51). (d) Riazi-
Sahhaf method is given by Eq. (42) in which the constants
for n-alkanes given in Table 2.6 should be used. In using this
method, if the given value is boiling point, Eq. (2.49) should be
used to calculate M from Tb. Then the predicted M will be used
to estimate other properties. In this method Pc is calculated
from Eq. (2.43). For parts a, b, c, g, and h, Zc is calculated from
its definition by Eq. (2.8). (e) Lee-Kesler method for M, To,
Pc, and Zc are given in Eqs. (2.54), (2.69), (2.70), and (2.103),
respectively. Vc should be back calculated through Eq. (2.8)
using Tc, Pc, and Zc. (f) Similarly for the Cavett method, Tc and
Pc are calculated from Eqs. (2.71) and (2.72), while Vc is back
calculated from Eq. (2.8) with Zc calculated from Eq. (2.104).
(g) The Twu methods are expressed by Eqs. (2.73)-(2.92) for
M, To, Pc, and Vc. Zc is calculated from Eq. (2.8). (h) The Winn
64 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

0.30
9 DIPPR Data
....... Eq. 2.103
..... Eq. 2.104
~ Eq. 2.42

84
--,.-.-:,
@

r.)
9 @ ~176

9 "..

020. ~ i t s I i , , . i I , , , i i , , , ,

5 l0 15 20 25
Carbon Number

FIG. 2.8---Estimation of critical compressibility factor of n-alkanes from various

methods.

m e t h o d for M, To, a n d Pc are given by Eqs. (2.93)-(2.95). I n
part h, Vc is calculated from the Hall-Yarborough t h r o u g h
Eq. (2.100) a n d Zc is calculated t h r o u g h Eq. (2.8). S u m m a r y
of results is given i n Table 2.11. No j u d g e m e n t can be m a d e
o n accuracy of these different methods t h r o u g h this single-
p o i n t evaluation. However, methods of R i a z i - S a h h a f (Part d)
a n d Twu (Part g) give the m o s t accurate results for this par-
ticular case. The reason is that these m e t h o d s have specific
relations for n-alkanes family a n d n-hexatriacontane is hydro-
c a r b o n from this family. I n addition, the values for the critical
properties from DIPPR [20] are estimated values rather t h a n
true experimental values. # used in Eq. (2.10) to derive corresponding correlations for the
2.5.4 Prediction of Acentric Factor
Acentric factor, w, is a defined p a r a m e t e r that is n o t directly
measurable. Accurate values of the acentric factor can be ob-
tained t h r o u g h accurate values of T~, Pc, a n d vapor pressure
with use of Eq. (2.10). Attempts to correlate co with p a r a m e -
ters such as Tb a n d SG all have failed. However, for homolo-
gous h y d r o c a r b o n groups the acentric factor can be related to
molecular weight as given by Eqs. (2.42) or (2.44). For other
c o m p o u n d s the acentric factor should be calculated t h r o u g h
its definition, i.e., Eq. (2.10), with the use of a correlation to
estimate vapor pressure. Use of a n accurate correlation for
vapor pressure would result i n a more accurate correlation
for the acentric factor. Methods of the calculation of the vapor
pressure are discussed in Chapter 7. There are three simple
correlations for the e s t i m a t i o n of vapor pressure that can be
acentric factor. These three methods are presented here.
2.5.4.1 Lee-Kesler M e t h o d
They proposed the following relations for the e s t i m a t i o n of
acentric factor based o n their proposed correlation for vapor
pressure [27].

TABLE 2.11--Prediction of critical properties of n-hexatriacontane from different methods a (Example 2.7).
M Tc, K APe,bar Vc, cma/mol Zc
Part Method(s) Est.** %D Est. %D Est. %D Est. %D Est. %D
Data from DIPPR [20] 507.0 --. 874.0 ... 6.8 .-- 2090.0 --. 0.196 -.-
a R-D: Eq. (2.38) 445.6 -12.1 885.8 1.3 7.3 7.4 1894.4 -9.3 0.188 -4.2
b API Methods 512.7 1.1 879.3 0.6 7.37 8.4 1 8 4 9 . 7 -11.5 0.205 4.6
c R - D (ext.): Eq. (2.46a) ...... 870.3 -0.4 5.54 -18.5 1964.7 -6.0 0.150 -23.3
d R-S: Eqs. 2.42 &2.43 506.9 0 871.8 -0.3 5.93 -12.8 1952.5 -6.6 0.16 -18.4
e L - K Methods 508.1 0.2 935.1 7.0 5.15 -24.3 2425.9 16.0 0 . 1 6 1 -18.0
f Cavett & Eq. (2.104) ...... 915.5 4.7 7.84 15.3 . . . . . . . . . . . .
g Twu 513.8 1.3 882.1 0.9 6.02 -11.4 2010.0 -3.8 0.165 -15.8
h Winn and H - Y 552.0 8.9 889.5 1.77 7.6 11.8 2362.9 13.1 0.243 24.0
aThe references for the methods are (a) R-D: Riazi-Daubert [28]; (b) API: Methods in the API-TDB[2]; (c) Extended Riazi-Dubert [65]; (d) Riazi-Sahhaf [31];
(e) Kesler-Lee[12] and Lee-Kesler[27]; (f) Cavett[26];Twu [31]; (h) Winn [25] and Hall-Yarborough[75]. Est.: Estimatedvalue.%D:% relativedeviationdefined
in Eq. (2.134).
 2. CHARACTERIZATION AND PROPERTIES OF PURE HYDROCARBONS 65

F o r Tbr < 0.8 (<C20 ~ M < 280) E d m i s t e r method:

In Pc/1.01325- 5.92714 + 6.09648/Tb~ + 1.28862 In Tb~-- 0.169347T~ (2.109) o ) = 0.5899 [ ~ x log - 1

-

15.2518 - 15.6875/Tbr - 13.4721 In Tbr+ 0.43577Tb6r

(2.105)
w h e r e Pc is in b a r a n d Tbr is the r e d u c e d boiling p o i n t w h i c h
is defined as
(2.106) Tbr = Tb/Tc
a n d Kesler-Lee [12] p r o p o s e d the following r e l a t i o n for Tbr >
0.8 (~>C20 ~ M > 280):
0) = - 7 . 9 0 4 + 0.1352Kw - 0.007465K 2 + 8.359Tb~
(2.107) + (1.408 -- O.Ol063Kw)/Tbr
in w h i c h Kw is the Watson c h a r a c t e r i z a t i o n factor defined by
Eq. (2.13). E q u a t i o n (2.105) m a y also be used for c o m p o u n d s
heavier t h a n C20 (Tbr > 0.8) w i t h o u t m a j o r e r r o r as s h o w n in
the e x a m p l e below
2.5.4.2 Edmister Method
The E d m i s t e r correlation [76] is developed o n the s a m e basis
as Eq. (2.105) b u t using a s i m p l e r t w o - p a r a m e t e r e q u a t i o n
for the v a p o r p r e s s u r e derived from Clapeyron e q u a t i o n (see
Eq. 7.15 in C h a p t e r 7).
(2.108) o)= (3)x (Tbr ~ X I1Ogl0 ( Pc
' tion (2.107) is a p p l i c a b l e for heavy fractions a n d a detailed
w h e r e logm0 is the l o g a r i t h m b a s e 10, Tbr is the r e d u c e d boiling
point, a n d Pc is the critical p r e s s u r e in bar. As is clear from
Eqs. (2.105) a n d (2.108), these two m e t h o d s require the s a m e
t h r e e i n p u t p a r a m e t e r s , namely, boiling point, critical temper-
ature, a n d critical pressure. E q u a t i o n s (2.105) a n d (2.108) are
directly derived f r o m v a p o r p r e s s u r e correlations discussed in
C h a p t e r 7.
2.5.4.3 Korsten Method
The E d m i s t e r m e t h o d u n d e r e s t i m a t e s acentric factor for
heavy c o m p o u n d s a n d the e r r o r tends to increase with in-
creasing m o l e c u l a r weight of c o m p o u n d s b e c a u s e the v a p o r
p r e s s u r e r a p i d l y decreases. M o s t recently K o r s t e n [77] m o d -
ified the Clapeyron e q u a t i o n for v a p o r p r e s s u r e of hydro-
c a r b o n systems a n d derived an e q u a t i o n very s i m i l a r to the
\ 1 - Tr r / 1.0~25
To c o m p a r e this e q u a t i o n with the E d m i s t e r equation, the
factor (3/7), w h i c h is equivalent to 0.42857 in Eq. (2.108), has
been r e p l a c e d by 0.58990 a n d the exponent of Tb~ has been
c h a n g e d from 1 to 1.3 in Eq. (2.109).
One c a n realize t h a t a c c u r a c y of these m e t h o d s m a i n l y de-
p e n d s on the a c c u r a c y of the i n p u t p a r a m e t e r s . However, for
p u r e c o m p o u n d s in w h i c h e x p e r i m e n t a l d a t a on p u r e hydro-
c a r b o n s are available the L e e - K e s l e r m e t h o d , Eq. (2.105),
gives a n AAD of 1-1.3%, while the E d m i s t e r m e t h o d gives
h i g h e r e r r o r of a b o u t 3-3.5%. The K o r s t e n m e t h o d is n e w
a n d it has not b e e n extensively evaluated for p e t r o l e u m frac-
tions, b u t for p u r e h y d r o c a r b o n s it seems t h a t it is m o r e ac-
curate t h a n the E d m i s t e r m e t h o d b u t less a c c u r a t e t h a n the
L e e - K e s l e r m e t h o d . Generally, the E d m i s t e r m e t h o d is not
r e c o m m e n d e d for p u r e h y d r o c a r b o n s a n d is u s e d to calcu-
late acentric factors of undefined p e t r o l e u m fractions. F o r
p e t r o l e u m fractions, the p s e u d o c r i t i c a l t e m p e r a t u r e a n d pres-
sure n e e d e d in Eqs. (2.105) a n d (2.108) m u s t be e s t i m a t e d
from m e t h o d s discussed in this section. Usually, w h e n the
Cavett o r W i n n m e t h o d s are used to estimate Tc a n d Pc,
the acentric factor is calculated b y the E d m i s t e r m e t h o d .
All o t h e r m e t h o d s for the e s t i m a t i o n of critical p r o p e r t i e s
use Eq. (2.105) for calculation of the acentric factor. Equa-
evaluation of its a c c u r a c y is n o t available in the literature. Fur-
ther evaluation of these m e t h o d s is given in Section 2.9. The
m e t h o d s of calculation of the acentric factor for p e t r o l e u m
fractions are discussed in the next chapter.
Example 2.8---Critical p r o p e r t i e s a n d acentric factor of
n-hexatriacontane (C36H74) are given as b y DIPPR [20] as
Tb ----770.2K, SG = 0.8172, Tc = 874.0 K, Pc = 6.8 bal; and~o =
1.52596. E s t i m a t e the acentric factor o f n - h e x a t r i a c o n t a n e us-
ing the following m e t h o d s :
a. Kesler-Lee m e t h o d with Tc, Pc from DIPPR
b. L e e - K e s l e r m e t h o d with To Pc from DIPPR
c. Edrnister m e t h o d with Tc, Pc from DIPPR
d. K o r s t e n m e t h o d with To Pc from DIPPR
e. R i a z i - S a h h a f correlation, Eq. (2.42)

TABLE 2.12--Prediction of acentric factor of n-hexatriacontane from different

methods (Example 2.8).
Method for %
Part Methodfor o) Tc & pa To. K Pc, bar Calc. o) Rel. de~
a Kesler-Lee DIPPR 874.0 6.8 1.351 -11.5
b Lee-Kesler DIPPR 874.0 6.8 1.869 22.4
c Edmister DIPPR 874.0 6.8 1.63 6.8
d Korsten DIPPR 874.0 6.8 1.731 13.5
e Riazi-Sahhaf not needed ...... 1.487 -2.6
f Korsten R-D-80 885.8 7.3 1.539 0.9
g Lee-Kesler API 879.3 7.4 1.846 21.0
h Korsten Ext. RD 870.3 5.54 1.529 0.2
i Lee-Kesler R-S 871.8 5.93 1.487 -2.6
j Edmister Winn 889.5 7.6 1.422 -6.8
k Kesler-Lee L-K 935.1 5.15 0.970 -36.4
1 Lee-Kesler Twu 882.1 6.03 1.475 -3.3
aR-D-80: Eqs. (2.63) and (2.64); API:Eqs. (2.65) and (2.66); Ext. RD: Eqs. (2.67) and (2.68);
R-S: Eqs. (2.42) and (2.43); Winn: Eqs. (2.94) and (2.95); L-K:Eqs. (2.69) and (2.70);
Twu: Eqs. (2.80) and (2.86).
66 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
f. Lee-Kesler method with To Pc obtained from Part a in
Example 2.6
g. Lee-Kesler method with Tc, Pc obtained from Part b in
Example 2.6
h. Lee-Kesler method with To Pc obtained from Part c in
Example 2.6
i. Lee-Kesler method with To Pc obtained from Part d in
Example 2.6
j. Edmister method with Tc, Pc obtained from Part h in Ex-
ample 2.6
k. Lee-Kesler method with Tc, Pc obtained from Part e in
Example 2.6
1. Lee-Kesler method with T~, Pc obtained from Part g in
Example 2.6
m. Tabulate %D for estimated value of acentric factor in each
method.
Lee-Kesler method refers to Eq. (2.105) and Kesler-Lee to
Eq. (2.107).
SolutionkAll three methods of Lee-Kesler, Edmister, and
Korsten require Tb, To, and Pc as input parameters. The
method of Kesler-Lee requires Kw in addition to Tbr. From
definition of Watson K, we get Kw --- 13.64. Substituting these
values from various methods one calculates the acentric fac-
tor. A summary of the results is given in Table 2.12. The least
accurate method is the Kesler-Lee correlations while the most
accurate method is Korsten combined with Eqs. (2.67) and
(2.68) for the critical constants, t (2.113)
2.6 P R E D I C T I O N OF DENSITY,
R E F R A C T I V E I N D E X , CH W E I G H T
RATIO, A N D F R E E Z I N G P O I N T
Estimation of density at different conditions of temperature,
pressure, and composition (p) is discussed in detail in Chap-
ter 5. However, liquid density at 20~ and 1 atm designated by
d in the unit of g/cm 3 is a useful characterization parameter
which will be used in Chapter 3 for the compositional analy-
sis of petroleum fractions especially in conjunction with the
definition of refractivity intercept by Eq. (2.14). The sodium
D line refractive index of liquid petroleum fractions at 20~
and 1 atm, n, is another useful characterization parameter.
Refractive index is needed in calculation of refractivity inter-
cept and is used in Eq. (2.40) for the estimation of various
properties through parameter I defined by Eq. (2.36). More-
over refractive index is useful in the calculation of density and
transport properties as discussed in Chapters 5 and 8. Carbon-
to-hydrogen weight ratio is needed in Chapter 3 for the esti-
mation of the composition of petroleum fractions. Freezing
point, TF, is useful for analyzing solidification of heavy com-
ponents in petroleum oils and to determine the cloud point
temperature of crude oils and reservoir fluids as discussed in
Chapter 9 (Section 9.3.3).
2.6.1 Prediction of Density at 20~
Numerical values of d20 for a given compound is very close
to the value of SG, which represents density at 15.5~ in the
unit of g/cm 3 as can be seen from Tables 2.1 and 2.3. Liquid
density generally decreases with temperature. Variation of
density with temperature is discussed in Chapter 6. However,
in this section methods of estimation of density at 20~ d20,
are presented to be used for the characterization methods dis-
cussed in Chapter 3. The most convenient way to estimate d20
is through specific gravity. As a rule of thumb d20 = 0.995 SG.
However, a better approximation is provided through calcu-
lation of change of density with temperature (Ad/AT), which
is negative and for hydrocarbon systems is given as [7]
(2.110) Ad/AT = - 1 0 -3 x (2.34 - 1.9dr)
where dr is density at temperature T in g/cm 3. This equation
may be used to obtain density at any temperature once a value
of density at one temperature is known. This equation is quite
accurate within a narrow temperature range limit. One can
use the above equation to obtain a value of density, d20, at
20~ (g/cm 3) from the specific gravity at 15.5~ as
(2.111) d20 = SG - 4.5 x 10-3(2.34 - 1.9SG)
Equation (2.111) may also be used to obtain SG from density
at 20 or 25~
SG -- 0.9915d20 + 0.01044
(2.112)
SG = 0.9823d25 + 0.02184
Similarly density at any other temperature may be calculated
through Eq. (2.110). Finally, Eq. (2.38) may also be used to
estimate d20 from TB and SG in the following form:
d20 = 0.983719Tb~176176 1~176
This equation was developed for hydrocarbons from C5 to C20;
however, it can be safely used up to C40 with AAD of less than
0.1%. A comparison is made between the above three meth-
ods of estimating d for some n-paraffins with actual data taken
from the API-TDB [2]. Results of evaluations are given in
Table 2.13. This summary evaluation shows that Eqs. (2.111)
and (2.113) are almost equivalent, while as expected the rule
of thumb is less accurate. Equation (2.111) is recommended
for practical calculations.
2.6.2 Prediction of Refractive Index
The refractive index of liquid hydrocarbons at 20~ is corre-
lated through parameter I defined by Eq. (2.14). If parameter
I is known, by rearranging Eq. (2.14), the refractive index, n,
can be calculated as follows:
(2.114) n = (11+~2//) 1/2
For pure and four different homologous hydrocarbon com-
pounds, parameter I is predicted from Eq. (2.42) using molec-
ular weight, M, with constants in Table 2.6. If boiling point is
available, M is first calculated by Eq. (2.48) and then I is cal-
culated. Prediction of I through Eq. (2.42) for various hydro-
carbon groups is shown in Fig. 2.9. Actual values of refractive
index from API-TDB [2] are also shown in this figure.
For all types of hydrocarbons and narrow-boiling range
petroleum fractions the simplest method to estimate param-
eter I is given by Riazi and Daubert [28] in the form of
Eq. (2.38) for the molecular weight range of 70-300 as follows:
(2.115) I = 0.3773Tb-~176 0~9182
 2. CHARACTERIZATION AND PROPERTIES OF PURE HYDROCARBONS 67

TABLE 2.13--Prediction of density (at 20~ of pure hydrocarbons.

Estimated density, g/cm3
n-Paraffin Tb, K SG d, g/cm3 Eq. (2.113) %AD Eq. (2.111) %AD 0.995SG %AD
n-C5 309.2 0.6317 0.6267 0.6271 0.06 0.6266 0.02 0.6285 0.29
n-Clo 447.3 0.7342 0.7303 0.7299 0.05 0.7299 0.05 0.7305 0.03
n-Cl5 543.8 0.7717 0.768 0.7677 0.03 0.7678 0.03 0.7678 0.02
n-C2o 616.9 0.7890 0.7871 0.7852 0.24 0.7852 0.24 0.7851 0.26
n-C25 683.2 0.8048 0.7996 0.8012 0.20 0.8012 0.19 0.8008 0.15
n-C3o 729.3 0.8123 0.8086 0.8088 0.03 0.8087 0.01 0.8082 0.04
n-C36 770.1 0.8172 0.8146 0.8138 0.09 0.8137 0.12 0.8131 0.18
Overall 0.10 0.10 0.14

w h e r e Tb is in Kelvin. This e q u a t i o n predicts n with a n aver- n u m b e r s greater t h a n C20. If o t h e r p a r a m e t e r s are available

age e r r o r of a b o u t 1% for p u r e h y d r o c a r b o n s from C5 to C20.
More a c c u r a t e relations are given b y Eq. (2.40) a n d Table 2.5
in t e r m s of various i n p u t p a r a m e t e r s . The following m e t h o d
developed b y Riazi a n d D a u b e r t [29] a n d i n c l u d e d in the API-
TDB [2] have a c c u r a c y of a b o u t 0.5% on n in the m o l e c u l a r
weight range of 70-300.
I = 2.34348 • 10 -2 [exp (7.029 • 10-4Tb + 2.468SG
(2.116) - 1.0267 x 10-3TbSG)] Tb~176 -0"720
w h e r e Tb is in kelvin. F o r heavier h y d r o c a r b o n s (>C20) the
following e q u a t i o n derived f r o m Eq. (2.46b) in t e r m s of M
a n d SG can be used.
I = 1.2419 x 10 -2 [exp (7.272 x 10-4M + 3.3223SG
(2.117) - 8 . 8 6 7 • 10-4MSG)] M~176176 -1"6117
E q u a t i o n (2.117) is generally a p p l i c a b l e to h y d r o c a r b o n s w i t h
a m o l e c u l a r weight range of 70-700 with an a c c u r a c y of less
t h a n 0.5%; however, it is m a i n l y r e c o m m e n d e d for c a r b o n
Eqs. (2.40) m a y be u s e d with constants given in Tables 2.5
a n d 2.9. The API m e t h o d to e s t i m a t e I for h y d r o c a r b o n s w i t h
M > 300 is s i m i l a r to Eq. (2.116) with different n u m e r i c a l
constants. Since for heavy fractions the boiling p o i n t is usu-
ally n o t available, Eq. (2.117) is p r e s e n t e d here. A n o t h e r rela-
tion for e s t i m a t i o n of I for heavy ihydrocarbons in t e r m s of Tb
a n d SG is given by Eq. (2.46a) with p a r a m e t e r s in Table 2.9,
w h i c h c a n be u s e d for heavy h y d r o c a r b o n s if distillation d a t a
is available.
Once refractive index at 20~ is estimated, the refractive
index at o t h e r t e m p e r a t u r e s m a y be p r e d i c t e d from the
following e m p i r i c a l relation [37].
(2.118) nr = n20 - 0.0004(T - 293.15)
where n20 is refractive index at 20~ (293 K) a n d n r is the
refractive index at the t e m p e r a t u r e T in w h i c h T is in kelvin.
Although this equation is simple, b u t it gives sufficient
a c c u r a c y for p r a c t i c a l applications. A m o r e a c c u r a t e relation
can be developed b y considering the slope of dnr/dT (value

1.6

n-alkylbenzenes
1.5

n-alkylcyclopentanes

4~

1.4

3 , , , , , ,,,I , , , , , ,,,, , , , , ,,,

10 100 1000 10000

Molecular Weight

FIG. 2.9--Prediction of refractive indices of pure hydrocarbons from Eq. (2.42).

68 CHARACTERIZATION AND PROPERTIES OF PETROLEUM
of -0.0004 in Eq. (2.118)) as a function of n20 rather than
a constant. Another approach to estimate refractive index
at temperatures other than 20~ is to assume that specific
refraction is constant for a given hydrocarbon:
lr I2o
(2.119) Specific refraction - dr - d2o - constant
where I2o is the refractive index parameter at 20~ and Ir is its
value at temperature T. Similarly dr is density at temperature
T. In fact the value of specific refraction is the same at all tem-
peratures [38]. If/20, d2o, and dT are known, then IT can be esti-
mated from the above equation. Value of nr can be calculated
from IT and Eq. (2.114). Equation (2.119) has the same accu-
racy as Eq. (2.118), but at the temperatures far from the refer-
ence temperature of 20 ~C accuracy of both methods decrease.
Because of simplicity, Eq. (2.118) is recommended for calcu-
lation of refractive index at different temperatures. It is obvi-
ous that the reference temperature in both Eqs. (2. I 18) and
(2.119) can be changed to any desired temperature in which
refractive index is available. Refractive index is also related to
another property called dielectric constant, e, which for non-
polar compounds at any temperature is e -- n 2. For example,
at temperature of 20~ a paraffinic oil has dielectric constant
of 2.195 and refractive index of 1.481 (n 2 -- 2.193). Dielectric
constants of petroleum products may be used to indicate the
presence of various constituents such as asphaltenes, resins,
etc. [11]. However, for more complex and polar molecules
such as muhiring aromatics, this simple relation between e
and n 2 is not valid and they are related through dipole mo-
ment. Further discussion on the methods of estimation of
refractive index is given by Riazi and Roomi [37].
2.6.3 Prediction of CH Weight Ratio
Carbon-to-hydrogen weight ratio as defined in Section 2.1.18
is indicative of the quality and type of hydrocarbons present
in a fuel. As will be shown in Chapter 3 from the knowledge
of CH value, composition of petroleum fractions may be es-
timated. CH value is also related to carbon residues as it is
discussed in the next chapter. For hydrocarbons with molec-
ular weight in the range of 70-300, the relations to estimate
CH values are given through Eq. (2.40) and Table 2.5. In terms
of TD and SG the relation is also given by Eq. (2.120) which
is also recommended for use in prediction of composition of
petroleum fractions [78].
CH = 3.4707 [exp (1.485 x 10-2Tb + 16.94SG
(2.120) -1.2492 x 10-2TbSG)] Tb2'725SG -6'798
where Tb is in kelvin. The above equation was used to extend
its application for hydrocarbons from C6 to C50.
CH = 8.7743 x 10 -l~ [exp (7.176 • 10-3Tb + 30.06242SG
(2.121) -7.35 x 10-3TbSG)] Tb~ -18"2753
where Tb is in kelvin. Although this equation was developed
based on data in the range of C20-C50, it can also be used
for lower hydrocarbons and it gives AAD of 2% for hydrocar-
bons from C20 to C50. Most of the data used in the develop-
ment of this equation are from n-alkanes and n-alkyl mono-
cyclic naphthenic and aromatic compounds. Estimation of
CH weight ratio from other input parameters is possible
through Eq. (2.40) and Table 2.5. Once CH weight ratio is
FRACTIONS
determined the atomic HC ratio can be calculated from their
definitions as described in Section 2.1.18:
11.9147
(2.122) HC (atomic ratio) =
CH(weight ratio)
E x a m p l e 2.9--Estimate the values of CH (weight) and HC
(atomic) ratios for n-tetradecylbenzene (C20H34) from Eqs.
(2.120) and (2.121) and compare with the actual value. Also
draw a graph of CH values from C6 to C50 for the three homol-
ogous hydrocarbon groups from paraffins, naphthenes, and
aromatics based on Eq. (2.121) and actual values.
Solution--The actual values of CH weight and HC
atomic ratios are calculated from the chemical formula
and Eq. (2.122) as CH --- (20 x 12.011)/(34 x 1.008) -- 7.01,
HC(atomic) = 34/20 = 1.7. From Table 2.1, for n-tetradecyl-
b e n z e n e (C20H34), Tb ~ 627 K and S G = 0.8587. Substitut-
ing these values into Eq. (2.120) gives CH-- 7.000, and from
Eq. (2.122) atomic HC ratio--1.702. The error from
Eq. (2.134) is %D = 0.12%. Equation (2.121) gives CH--
6.998, which is nearly the same as the value obtained from
Eq. (2.120) with the same error. Similarly CH values are cal-
culated by Eq. (2.121) for hydrocarbons ranging from C6 to
C50 in three homologous hydrocarbon groups and are shown
with actual values in Fig. 2.10. #
2.6.4 Prediction of Freezing/Melting Point
For pure compounds, the normal freezing point is the same
as the melting point, TM. Melting point is mainly a param-
eter that is needed for predicting solid-liquid phase behav-
ior, especially for the waxy oils as shown in Chapter 9. All
attempts to develop a generalized correlation for TM in the
form of Eq. (2.38) have failed. However, Eq. (2.42) developed
by Riazi and Sahhaf for various homologous hydrocarbon
groups can be used to estimate melting or freezing point of
pure hydrocarbons from C7 to C40 with good accuracy (error
of 1-1.5%) for practical calculations [31]. Using this equation
with appropriate constants in Table 2.6 gives the following
equations for predicting the freezing point of n-alkanes (P),
n-alkycyclopentanes (N), and n-alkybenzenes (A) from molec-
ular weight.
(2.123) TMp = 397 -- exp(6.5096 -- 0.14187M ~
(2.124) TMN = 370 exp(6.52504-- 0.04945M 2/3)
(2.125) TMA= 395 -- exp(6.53599 -- 0.04912M 2/3)
where TM is in kelvin. These equations are valid in the car-
bon ranges of C5-C40, C7-C40, and C9-C40 for the P, N, and
A groups, respectively. In fact in wax precipitation linear
hydrocarbons from CI to CI5 as well as aromatics are ab-
sent, therefore there is no need for the melting point of very
light hydrocarbons [64]. Equation (2.124) is for the melt-
ing point of n-alkylcyclopentanes. A similar correlation for
n-alkylcyclohexanes is given by Eq. (2.42) with constants in
Table 2.6. In Chapter 3, these correlations will be used to es-
timate freezing point of petroleum fractions.
Won [79] and Pan et al. [63] also proposed correlations
for the freezing points of hydrocarbon groups. The Won
 2. CHARACTERIZATION AND PROPERTIES OF PURE HYDROCARBONS 69

10.0

t * Actual Values forn-alkanes

9 Actual Values forn-alkyicyclopentanes
0
.=~ 9.0 A Actual Values forn-alkybcnzenes

O8

8.0
O8

7.0
0

6,0 .~;~..o e l , ~ eo ~ ooeo o l e o o e o , I . Q , ooo~ o o e o o e ~ 9

5.0
5 15 25 35 45 55
Carbon Number

FIG. 2.10--Estimation of CH weight Ratio from Eq. (2.121) for various families.

50

0
. Y..P-

-50
o Data for n-alkanes
Predicted: R-S Method
N
Predicted: P-F Method
~ -100 A Data for n-alkyl~'clopenlanes
. . . . . . . Predicted: R-S Method
,, K
~
iI .... Predicted: P-F Method
It I Data for n-alkylbenzenes
-150 - - - - - Predicted: R-S Method
m Predicted: P-F Method

-200
5 10 15 20 25
Carbon Number

FIG. 2.11--Estimation of freezing point of pure hydrocarbons for various fami-

lies. [R-S refers to Eqs. (2.123)-(2.125); P-F refers to Eqs. (2.126) and (2.127).
70 CHARACTERIZATION AND P R O P E R T I E S OF P E T R O L E U M FRACTIONS
correlation for n-alkanes is
(2.126) Trap = 374.5 + 0.02617M - 20172/M
where TMe is in kelvin. For naphthenes, aromatics, and
isoparaffins the melting point temperature may be esti-
mated from the following relation given by Pan-Firrozabadi-
Fotland [63].
(2.127) Tra(iP,N,A)= 333.45 -- 419 exp(-0.00855M)
where TM is in kelvin. Subscripts iP, N, and A indicate iso-
paraffins, naphthenes, and aromatics, respectively.
Example 2.10--Estimate the freezing point of n-hexa-
triacontane (C36H74 ) from Eqs. (2.123 ) and (2.126 ) and com-
pare with the actual value of 348.19 K [20]. Also draw a graph
of predicted Tra from Eqs. (2. i23) to (2.127) for hydrocarbons
from C7 to C40 for the three homologous hydrocarbon groups
from paraffins, naphthenes, and aromatics based on the above
two methods and compare with actual values given up to C20
given in Table 2.2.
Solution--For n-C36, we have M = 36 x 12.011 + 74 x 1.008 =
508.98 and T~a = 348.19 K. From Eq. (2.123), TM = 3 9 7 -
exp(6.5096- 0.14187 • 508.980'47) ----349.78 K. The percent
absolute relative deviation (%AD) is 0.2%. Using Eq. (2.126),
T~a = 348.19 K with %AD of 0.24%. A complete evaluation is
demonstrated in Fig. 2.1 I. On an overall basis for n-alkanes
Eq. (2.126) is more accurate than Eq. (2.123) while for naph-
thenes and aromatics, Eqs. (2.124) and (2.125) are more ac-
curate than Eq. (2.127).
2.7 P R E D I C T I O N O F KINEMATIC
V I S C O S I T Y AT 38 A N D 9 9 ~
Detailed prediction of the viscosities of petroleum fractions
will be discussed in Chapter 8. However, kinematic viscos-
ity defined by Eq. (2.12) is a characterization parameter
needed to calculate parameters such as VGC (Section 2.1.17),
which will be used in Chapter 3 to determine the compo-
sition of petroleum fractions. Kinematic viscosity at two
reference temperatures of 100~ (37.78 ~ 38~ and 210~
(98.89 ~ 99~ are generally used as basic characterization
parameters and are designated by 1)38(100)and 1)99(210),respec-
tively. For simplicity in writing, the reference temperatures of
100 and 210~ are presented as 38 and 99~ rather than accu-
rate values of 37.78 and 98.89. Kinematic viscosity decreases
with temperature and for highly viscous oils values of 1)99(210)
are reported rather than u38~00). The temperature dependency
of viscosity is discussed in Chapter 8 and as will be seen, the
viscosity of petroleum fractions is one of the most complex
physical properties to predict, especially for very heavy frac-
tions and multiring aromatic/naphthenic compounds. Heavy
oils with API gravities less than 10 could have kinematic vis-
cosities of several millions cSt at 99~ (210~ These viscos-
ity values would be almost impossible to predict from bulk
properties such as boiling point and specific gravity. How-
ever, there are some relations proposed in the literature for
the estimation of these kinematic viscosities from Tb and SG
or their equivalent parameters Kw and API gravity. Relations
developed by Abbott et al. [80] are commonly used for the
estimation of reference kinematic viscosities and are also in-
cluded in the API-TDB [2]:
log v38000) =
4.39371 - 1.94733Kw + 0.12769K~v
+3,2629 x 10-4API 2 - 1.18246 x 10-ZKwAPI
0.171617K2w + 10.9943(API) + 9.50663 x 10-2(API) 2 - 0.860218Kw(API)
+
(API) + 50.3642 -4.78231 Kw
(2.128)
log 1)99(210) =
- 0 . 4 6 3 6 3 4 - 0.166532(API) + 5.13447
• 10-4(API)2 - 8.48995 x 10-3KwAPI
8.0325 • 10-2 Kw + 1.24899(API) + 0.19768(API) 2
+
(API) + 26.786 - 2.6296Kw
(2.129)
Kw and API are defined by Eqs. (2.13) and (2.4). In these
relations the kinematic viscosities are in cSt (mm2/s). These
correlations are also shown by a nomograph in Fig. 2.12. The
above relations cannot be applied to heavy oils and should
be used with special care when Kw < 10 or Kw > 12.5 and
API < 0 or API > 80. Average error for these equations is in
the range of 15-20%. They are best applicable for the viscos-
ity ranges of 0.5 < 1)38(100) < 20 mm2/s and 0.3 < 1)99(210)< 40
mm2/s [8]. There are some other methods available in the lit-
erature for the estimation of kinematic viscosities at 38 and
99~ For example Twu [81 ] proposed two correlations for the
kinematic viscosities of n-alkanes from C1 to C100in a similar
fashion as his correlations for the critical properties discussed
in Section 2.5.1. Errors of 4-100% are common for prediction
of viscosities of typical oils through this method [ 17].
Once kinematic viscosities at two temperatures are known,
ASTM charts (ASTM D 341-93) may be used to obtain viscos-
ity at other temperatures. The ASTM chart is an empirical
relation between kinematic viscosity and temperature and it
is given in Fig. 2.13 [68]. In using this chart two points whose
their viscosity and temperature are known are located and
a straight line should connect these two points. At any other
temperature viscosity can be read from the chart. Estimated
values are more accurate within a smaller temperature range.
This graph can be represented by the following correlation
[8]:
(2.130) log[log(1)r + 0.7 + c:~)] = A1 + B1 log T
where vr is in cSt, T is the absolute temperature in kelvin,
and log is the logarithm to base 10. Parameter Cr varies with
value of Vr as follows [8]:
0.085(1)r - 1.5) 2 if vr < 1.5 cSt [mm2/s]
(2.131) Cr =
0.0 if 1)~ >_ 1.5 cSt [mmZ/s]
If the reference temperatures are 100 and 210~ (38 and
99~ then A1 and B1 are given by the following relations:
Al -----12.8356 x (2.57059D1 - 2.49268D2)
B1 = 12.8356(D2 - D1)
(2.132)
D1 = log[log(v38000)+ 0.7 + c38000))]
/)2 -----log[log(1)99(210)+ 0.7 + C99(210)) ]
Various forms of Eq. (2.130) are given in other sources
[2, 11, 17]. Errors arising from use of Eq. (2.130) are better or
 2. C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P U R E H Y D R O C A R B O N S 71

12.5--

165

50

IZO-

1.0~ 40

I/I

1.2

0 x/
u2 ~" 11..5"
30 O

t-
1.6- O

0
g N
1.8
u~ 0 0_
<(
t-
a E
9 ~ O
t-
.~o O
O) 20
0 ~ .! 1.0"
u~
0

4 3o

0
!0

10.5 -
s--t L-2o0

I soo
1 7oo

70~

lO.O-J--

FIG. 2.12--Prediction of kinematic viscosity from Kw and the API gravity. With
permission from Ref. [2],

at least in the same range of errors for the prediction of viscos-
ity from Eqs. (2.128) and (2.129). Similarly constants AI and
B1 in Eq. (2.130) can be determined when values of viscosity
at two temperatures other than 100 and 210~ are known.
When vr is being calculated from Eq. (2.130) at temperature
T, a trial and error procedure is required to determine param-
eter cr. The first estimate is calculated by assuming vr > 1.5
cSt and thus Cr = 0. If calculated value is less than 1.5 cSt,
then Cr is calculated from Eq. (2.131). Extrapolated values
from Fig. 2.12 or Eq. (2.130) should be taken with caution.
An application of this method to estimate kinematic viscosity
of petroleum fractions is demonstrated in Chapter 3. Further
discussion on the estimation of viscosity is given in Chapter 8.
Consistency Text--One way to check reliability of a predicted
physical property is to perform a consistency test through
different procedures. For example, laboratory reports may
consist of viscosity data at a temperature other than 38 or
 Temperature, Degrees Celsius

u)

O
(D
t-
(9 3
C~

u) <
O (..)
>
O ~z
03 (l)
E
.c_
v 0
~0
Temperature, Degrees Celsius

FIG. 2.13mViscosity-temperature relation: ASTM chart [68].

 2. CHARACTERIZATION AND PROPERTIES OF PURE HYDROCARBONS 73
99~ In many cases kinematic viscosity at 40~ (122~ or
60~ (140~ is reported. One may estimate the kinematic
viscosities at 38 and 99~ through Eqs. (2.128) and (2.129)
and then use the ASTM chart (or Eq. 2.128) to obtain the
value of viscosity at 40 or 60~ If the interpolated value is far
from laboratory data then the estimation method cannot be
trusted and other methods should be considered.
Another consistency test can be made through estimation
of the molecular weight by Eq. (2.52) using estimated vis-
cosities by Eqs. (2.128) and (2.129). If value of M calculated
through Eq. (2.52) is near the value of M estimated from Tb
and SG by Eqs. (2.51) or (2.50), then all estimated values can
be trusted. Such consistency tests can be extended to all other
physical properties. The following example demonstrates the
test method.
Example 2.11--The viscosity of a pure multiring hydrocar-
bon from an aromatic group (naphthecene type compound)
with formula C 2 6 H 4 0 has been i~eported in the API RP-42
[ 18]. Data available are M = 352.6, SG = 0.9845, and 1299(210) =
13.09 cSt. Estimate the kinematic viscosity of this hydrocar-
bon at 38 and 99~ (100 and 210~ by Eqs. (2.128) and
(2.129). How can you assess the validity of your estimated
kinematic viscosity at 38~
Solution--To estimate the viscosity through the Abbott cor-
relations, Kw and API gravity are needed. However, Tb is not
available and should be estimated from M and SG. Since
M > 300, we use Eq. (2.57) in terms of M and SG to esti-
mate Tb as follows: Tb = 720.7 K. Using Eqs. (2.13) and (2.4),
Kw and API are calculated as Kw = 11.08 and API = 12.23.
Using Eqs. (2.128) and (2.129) the viscosities are calculated
as 1)38(100) = 299.76 cSt. 1399(210) ~ I 1.35 cSt. At 99~ the esti-
mated value can be directly evaluated against the experimen-
tal data: %D = [(11.08 - 13.09)/13.09)] x 100 = -15.4%. To
evaluate accuracy of estimated viscosity at 38~ a consistency
test is required. Since the actual value of molecular weight,
M, is given, one can estimate M through Eq. (2.52) using
estimated values of 1)38(100), 1)99(210), and SG as the input pa-
rameter. The estimated M is 333.8 which in comparison with
actual value of 352.6 gives %AD of 11.36%. This error is ac-
ceptable considering that Eq. (2.52) has been developed based
on data of petroleum fractions and the fact that input param-
eters are estimated rather than actual values. Therefore, we
can conclude that the consistency test has been successful
and the value of 299.8 cSt as viscosity of this hydrocarbon at
38~ is acceptable. The error on estimated viscosity at 99~
is 15.4%, which is within the range of errors reported for the
method. It should be realized that the equations for prediction
of kinematic viscosity and estimation of molecular weight by
Eq. (2.52) were originally recommended for petroleum frac-
tions rather than pure compounds.
2.8 T H E W I N N N O M O G R A M
Development of estimation techniques through graphical
methods was quite common in the 1930s through the 1950s
when computational tools were not available. Nomogram or
homograph usually refers to a graphical correlation between
different input parameters and desired property when more
than two input parameters are involved. By drawing a straight
line between values of input parameters, a reading can be
made where the straight line intersects with the line (or curve)
of the desired property. The best example and widely used
nomogram is the one developed by Winn in 1957 [25]. This
nomogram, which is also included in the API-TDB [2], re-
lates molecular weight (M), CH weight ratio, aniline point,
and Watson K to boiling point and specific gravity (or API
gravity) on a single chart and is shown in Fig. 2.14.
Application of this figure is mainly for petroleum fractions
and the mean average boiling point defined in Chapter 3
is used as the boiling point, Tb. If any two parameters are
available, all other characterization parameters can be deter-
mined. However, on the figure, the best two input parameters
are Tb and SG that are on the opposite side of the figure.
Obviously use of only M and Tb as input parameters is not
suitable since they are near each other on the figure and an
accurate reading for other parameters would not be possible.
Similarly CH and SG are not suitable as the only two input
parameters. Previously the computerized form of the Winn
nomogram for molecular weight was given by Eq. (2.95).
Use of the nomogram is not common at the present time
especially with availability of personal computers (PCs) and
simulators, but still some process engineers prefer to use a
nomogram to have a quick estimate of a property or to check
their calculations from analytical correlations and computer
programs.
If the boiling point is not available, methods discussed in
Section 2.4.2 may be used to estimate the boiling point be-
fore using the figure. Equation (2.50) for molecular weight
may be combined with Eq. (2.13) to obtain a relation for the
estimation of Kw from M and SG [51].
(2.133) Kw = 4 . 5 5 7 9 M ~ -0"84573
This equation gives an approximate value for Kw and should
be used with care for hydrocarbons heavier than C30. A more
accurate correlation for estimation of Kw can be obtained if
the boiling point is calculated from Eq. (2.56) or (2.57) and
used in Eq. (2.13) to calculate Kw.
Example 2.12--Basic properties of n-tridecylcyclohexane
(CIgH38) a r e given in Table 2.1. Use M and SG as available
input parameters to calculate
a. Kw from Eq. (2.133).
b. Kw from most accurate method.
c. Kw from Winn Nomogram.
d. CH weight ratio from M and SG.
e. CH weight ratio from Winn method.
f. %D for each method in comparison with the actual values.
Solution--From Table 2.1, M = 266.5, Tb = 614.7 K,
#
and SG = 0.8277. HC atomic ratio -- 38/19 = 2.0. Using
Eq. (2.122), CH weight ratio = 11.9147/2.0 = 5.957. From
definition of Kw, i.e., Eq. (2.13), the actual value of Kw is
calculated as Kw = (1.8 x 614.7)1/3/0.8277 = 12.496.
a. From Eq. (2.133), Kw = 4.5579 x [(266.5) ~ x
[(0.8277) -~ = 12.485. %D = -0.09%.
b. The compound is from the n-alkylcyclohexane family and
the most accurate way of predicting its boiling point is
through Eq. (2.42) with constants given in Table 2.6 which
74 CHARACTERIZATIONAND PROPERTIES OF PETROLEUMFRACTIONS
90-i_ _--1200
L
ss --- -~s
:l 10O
8o-
-~ooo
75-2

70-_ -0.70

65" ~00 r

F-
L800
500
60-
- -0.75 O

55-- ~too
u.
"
240~ ~ ~00 70O

50-
.~ 22o..2 r .

c~
,3- '
5-300 -600 ~.

4s- -'oso '-' v 200- ~- #_

._.g,
0
40-
09 ~ ~
~,2.." ~
o
.
150--~ ~
~
"- 200
.

Lsoo~~
.-

,~
E ID
< 35- -0.85
.._ c-

.~ 150
3o-2
-- ~
2s- -o.9o ~a- ~-~ :_
:I: -
- -300

- o9-2 '
--0.95 -9o
15-
-80 -20O

10 tOO

5-

~t05
- -100
0-

FIG. 2.14~Winn nomogram for characterization of petroleum fractions. With

p e r m i s s i o n from Ref. [2].

gives Tb = 1100 -- exp[7.00275 -- 0.01977 x (266.5) 2/3] =
615.08 K. Using Tb = 615.08 a n d SG = 0.8277 in Eq. (2.13)
gives Kw = 12.498. The %D is +0.016%.
c. W h e n using Winn n o m o g r a m (Fig. 2.14) it is easier to con-
vert SG to API gravity, w h i c h t h r o u g h Eq. (2.4) is 39.46.
A straight line b e t w e e n points 266.5 on the M line an d
39.5 on the API gravity line intersects the Watson K line
at Kw = 12.27 an d the intersection with the CH line is at
CH = 6.1. The %D for Kw is - 1 . 8 % .
d. CH weight ratio can be e s ti m a t e d f r o m Eq. (2.40) using
M and SG as i n p ut p a r a m e t e r s with constants in Table 2.5,
w h i c h result in CH value of 6.2 with % D = + 4% ( C H a c t u a i =
5.96).
CH = 2.35475 x [exp(9.3485 • 10 -3 • 266.5 + 4.74695
x 0.8277 - 8.01719 x 10 -3 x 266.5 x 0.8277)]
x [(266,5) -0'68418] x [(0.8277) -0'7682] = 6.2
e, CH weight ratio f r o m the Winn m e t h o d was o b t a i n e d in
Part c as C H = 6 . 1 , w h i c h gives %D = +2.3%.
f. F o r all parts % D is calculated f r o m Eq. (2.134). Part b
gives the m o s t a c c u r a t e Kw value b ecau se Tb w a s calculated
 2. CHARACTERIZATION AND PROPERTIES OF PURE HYDROCARBONS
accurately. However, Part b in this case is also accurate. Es-
timation of the CH value is less accurate than prediction of
boiling point and gives errors higher than Kw.
2.9 ANALYSIS AND COMPARISON OF
VARIOUS CHARACTERIZATION METHODS
Generally there are a large number of pure hydrocarbons
and their properties can be used for evaluation purposes.
However, hydrocarbons from certain groups (i.e., paraffins,
naphthenes, and aromatics) are more abundant in petroleum
fractions and can be used as a database for evaluation pur-
poses. Molecular weight, critical properties, and acentric fac-
tor are important properties and their predictive methods
are presented in this chapter. Errors in any of these proper-
ties greatly influence the accuracy of the estimated physical
property. Methods of estimation of these properties from bulk
properties such as boiling point and specific gravity that are
presented in this chapter have been in use in the petroleum
industry for many years. In some process simulators a user
should select a characterization method out of more than a
dozen methods included in the simulator [56]. In each ap-
plication, the choice of characterization method by the user
strongly influences the simulation results. Although there has
not been a general and comprehensive evaluation of various
characterization methods, a conclusion can be made from
individual's experiences reported in the literature. In this sec-
tion first we discuss criteria for evaluation of various methods
and then different predictive methods for molecular weight,
critical constants and acentric factor are compared and eval-
uated.
2.9.1 Criteria for Evaluation
o f a Characterization M e t h o d
Methods of characterization and correlations presented in
this chapter are mainly based on properties of pure hydrocar-
bons. However, some of these correlations such as Eq. (2.52)
for estimation of the molecular weight of heavy fractions or
the correlations presented for prediction of the kinematic vis-
cosity are based mainly on the properties of fractions rather
than pure compounds. The main application of these correla-
tions is for basic properties of undefined petroleum fractions
in which bulk properties of a fraction are used to estimate
a desired parameter. Therefore, the true evaluation of these
characterization methods should be made through properties
of petroleum fractions as will be discussed in upcoming chap-
ters. However, evaluation of these methods with properties of
pure hydrocarbons can be used as a preliminary criteria to
judge the accuracy of various correlations. A method, which
is more accurate than other methods for pure hydrocarbons,
is not necessarily the best method for petroleum mixtures. A
database for pure hydrocarbons consists of many compounds
from different families. However, evaluations made by some
researchers are based primarily on properties of limited pure
hydrocarbons (e.g., n-alkanes). The conclusions through such
evaluations cannot be generalized to all hydrocarbons and
petroleum fractions. Perhaps it is not a fair comparison if
a data set used to develop a method is also used to evalu-
ate the other methods that have used other databases. Type
75
of compounds selected, the source of data, number of data
points, and the basis for the evaluation all affect evaluation
outcome. The number of numerical constants in a correla-
tion and number of input parameters also affect the accu-
racy. Usually older methods are based on a fewer and less
accurate data than newer methods. It would be always useful
to test different methods on a set of data that have not been
used in obtaining the correlation coefficients. The roost ap-
propriate procedure would be to compare various methods
with an independent data set not used in the development of
any methods considered in the evaluation process. Another
fair comparison of two different correlation would be to use
the same database and reobtain the numerical constants in
each correlation from a single database. This was done when
Eqs. (2.52) and (2.53) were compared, as discussed in Sec-
tion 2.4.1. These are the bases that have been used to compare
some of the correlations presented in this chapter.
Basically there are two parameters for the evaluation of
a correlation. One parameter is the percent average absolute
deviation (%AAD). Average errors reported in this chapter and
throughout the book are based on percent relative deviation
(%D). These errors are defined as following:
estimated value - actual value)
(2.134) %D = \ ac~u~ualva~ue x 100
where N is the total number of data points and summation is
made on all the points. I%D[ is called percent absolute devi-
ation and it is shown by %AD. The maximum value of I%D[
in a data set is referred as %MAD. The second parameter is
called R squared (R2) that is considered as an index of the cor-
relation when parameters of a correlation are obtained from
a data set. A value of 1 means perfect fit while values above
0.99 generally give good correlation. For a set of data with
X column (independent variable) and Y column (dependent
variable) the parameter is defined as
(2.136) R2= [N(y~xY)-(~X)(Y~Y)]2
[ N y ~ X 2 - ( ~ X ) 2] x [ N ~ Y 2 - (y~y)2]
where X and Y are values of the independent and correspond-
ing dependent variables and N is the number of data points.
The ~ is the summation over all N values of X, X 2, Y, y2, and
XY as indicated in the above equation. The R2 value can be
interpreted as the proportion of the variance in y attributable
to the variance in x and it varies from 0 to maximum value
of 1.
For most of the correlations presented in this chapter such
as Eqs. (2.40), (2.42), or (2.46a) the %AAD for various prop-
erties is usually given in the corresponding tables where the
constants are shown. Most of these properties have been cor-
related with an R2 value of minimum 0.99. Some of these
properties such as kinematic viscosity or CH weight ratio
showed lower values for R2. Evaluation of some of the other
correlations is made through various examples presented in
this chapter.
Nowadays with access to sophisticated mathematical tools,
it is possible to obtain a very accurate correlation from any
76 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

data set. For example, when the method of neural network is TABLE 2.14--Evaluation of methods for estimation of molecular
used to obtain correlations for estimation of critical proper- weight of petroleum fractions,a
Abs Dev%**
ties, a very accurate correlation can be obtained for a large
number of compounds [82]. However, such correlations con- Method Equation(s) AAD% MAD%
tain as m a n y as 30 numerical values, which limit their power API (Riazi-Daubert) (2.51 ) 3.9 18.7
of extrapolatability. It is our experience that when a corre- Twu (2.89)-(2.92) 5.0 16.1
Kesler-Lee (2.54) 8.2 28.2
lation is based on some theoretical foundation, it has fewer Winn (2.93) 5.4 25.9
constants with a wider range of application and better ex- ~ N u r n b e r o f d a t a p o i n t s : 625; R a n g e s o f d a t a : M ~ 7 0 - 7 0 0 , Tb ~ 3 0 0 - 8 5 0 , S G
trapolatability. This is particularly evident for the case of 0.63-0.97
b D e f i n e d b y E q s . (2.134) a n d (135). R e f e r e n c e [29].
Eq. (2.38) developed based on the theory of intermolecular
forces and EOS parameters. Equation (2.38) has only three
parameters that are obtained from data on properties of pure fractions. This equation has been included in most process
hydrocarbons from C5 to C20. This equation for various prop- simulators [54-56]. Whitson [51, 53] has used this equation
erties can be safely used up to C30. Tsonopoulos et aL [34] and its conversion to Kw (Eq. 2.133) for fractions up to C25
and Lin et al. [83] have extensively evaluated Eq. (2.50) for in his characterization methods of reservoir fluids. A more
estimation of the molecular weight of different samples of general form of this equation is given by Eq. (2.51) for the
coal liquids, which are mainly aromatics, and compared with molecular weight range of 70-700. This equation gives an
other sophisticated multiparameter correlations specifically average error of 3.4% for fractions with M < 300 and 4.7%
developed for the molecular weight of coal liquids. Their con- for fractions with M > 300 for 625 fractions from Penn State
clusion was that Eq. (2.50) gave the lowest error even though database on petroleum fractions. An advantage of Eq. (2.51)
only pure component data were used to develop this equation. over Eq. (2.50) is that it is applicable to both light and heavy
Further evaluation of characterization methods for molecu- fractions. A comparative evaluation of various correlations
lar weight and critical properties are given in the following for estimation of molecular weight is given in Table 2.14 [29].
parts. Process simulators [55] usually have referred to Eq. (2.50)
as Riazi-Daubert method and Eq. (2.51) as the API method.
The Winn method, Eq. (2.93), has been also referred as S i m -
2.9.2 Evaluation o f Methods o f Estimation
Daubert method in some sources [55, 84].
o f Molecular Weight
For pure hydrocarbons the molecular weight of three ho-
As mentioned above most of the evaluations made on mologous hydrocarbon groups predicted from Eq. (2.51) is
Eq. (2.50) for the molecular weight of petroleum fractions drawn versus carbon number in Fig. 2.15. For a given car-
below 300 suggest that it predicts quite well for various bon number the difference between molecular weights of

3000
o A P I Data
..... n-Alkanes
2500 Naphthenes
- - n-Alk3,1benzenes

2000 /
Y~
e~ /

1500
o

1000

500

I i f i I i f
0
200 400 600 800 1000 1200

Boiling Point, K
FIG. 2.15--Evaluation of Eq. (2.51) for molecular weight of pure compounds.
 2. CHARACTERIZATION AND P R O P E R TIE S OF PURE H YD R O CA R B O N S 77

400
o API D a t a ..M~" ,
-- R-D:Eq. 2.50 ~ A~."~
350
. . . . API: Eq. 2.51 ~ r
. . . . . . R.s: 248 /. G..?::
-- - --Twu: Eq. 2.89-2.92 ~ 7
300 Lee-Kcsler:E q ~

250

200 "'"
O
E
150

100

50 , , .., ~ , , , , r , , , , f , , , t , , , ,

5 10 15 20 25 30
Carbon Number
FIG. 2.16--Evaluation of various methods for prediction o1 molecular weight
of n-alkylcycohexanes. Riazi-Daubert: Eq. (2,50); API: Eq. (2.51); Riazi-Sahhaf:
Eq. (2.48); Lee-Kesler: Eq. (2.54); Twu: Eqs, (2,89)-(2,92).

hydrocarbons from different groups is small. Actual values of
molecular weight of n-alkylbenzenes up to C20 as reported by
API-TDB [2] are also shown on the figure. Equation (2.51) is
not the best method for the prediction of molecular weight of
pure compounds as it was primarily developed for petroleum
fractions. Various methods for the estimation of molecular
weight for n-alkylcylohexanes with the API data (up to C26)
are shown in Fig. 2.16 for the range of C6-Cs0. At higher car-
bon numbers the deviation between the methods increases.
The Twu method accurately estimates molecular weight of
low-molecular-weight pure hydrocarbons; however, at higher
molecular weights it deviates from actual data. A compari-
son between evaluations presented in Fig. 2.16 and Table 2.14
shows that a method that is accurate for prediction of proper-
ties of pure hydrocarbons is not necessarily the best method
for petroleum fractions. Evaluation of method of prediction
of molecular weight from viscosity (Eqs. (2.52) and (2.53))
has been discussed in Section 2.4.1.
2.9.3 Evaluation o f M e t h o d s of E s t i m a t i o n
of Critical Properties
Evaluation of correlations for estimation of critical proper-
ties of pure compounds can be made directly with the actual
values for hydrocarbons up to C18. However, when they are
applied to petroleum fractions, pseudocritical properties are
calculated which are not directly measurable. These values
should be evaluated through other properties that are mea-
surable but require critical properties for their calculations.
For example, enthalpies of petroleum fractions are calcu-
lated through generalized correlations which require critical
properties as shown in Chapters 6 and 7. The phase behav-
ior prediction of reservoir fluids also requires critical prop-
erties of petroleum cuts that make up the fluid as discussed
in Chapter 9. These two indirect methods are the basis for
the evaluation of correlations for estimation of critical prop-
erties. These evaluations very much depend on the type of
fractions evaluated. For example, Eqs. (2.63)-(2.66) for esti-
mation of Tc and Pc have been developed based on the critical
data from C5 to Cls; therefore, their application to heavy frac-
tions is not reliable although they can be safely extrapolated to
C25-C30 hydrocarbons. In the development of these equations,
the internal consistency between Tc and Pc was not imposed as
the correlations were developed for fractions with M < 300.
These correlations were primarily developed for light frac-
tions and medium distillates that are produced from atmo-
spheric distillation columns.
For pure hydrocarbons from homologous families,
Eq. (2.42) with constants in Table (2.6) provide accurate val-
ues for Tc, Pc, and Vc. Prediction of Tc and Pc from this equa-
tion and comparison with the API-TDB data are shown in
Figs. 2.2 and 2.3, respectively. Evaluation of various methods
for critical temperature, pressure, and volume of different hy-
drocarbon families is demonstrated in Figs. 2.17-2.19 respec-
tively. A summary of evaluations for Tc and Pc of hydrocarbons
from different groups of all types is presented in Table 2.15
[29]. Discontinuity of API data on Pc of n-alkylcyclopentanes,
as seen in Fig. 2.18, is due to prediction of Pc for heav-
ier hydrocarbons (>C20) through a group contribution
method.
78 CHARACTERIZATION AND P R O P E R TI E S OF PETROLEUM FRACTIONS

1050

o API Data

~ Winn

- - R - D _ , ~ . ~ , . ~. . ~. 9
- - - Twtl . . . ~ ~
900
~Z

E
750

600

450 l I I I I L I L L e i i i i i i i l i i i i i i i i i l

10 20 30 40 50 60
Carbon Number

FIG. 2.17--Comparison of various methods for estimation of critical temperature of n-

alkanes. API Data: API-TDB [2]; Winn: Eq. (2.94); R-D: Riazi-Daubert, Eq. (2.63); Twu:
Eq. (2.80)-(2.82); Ext. R-D: Extended Riazi-Daubert, Eq. (2.67); L-K: Lee-Kesler: Eq. (2.69);
API: Eq. (2.55); R-S: Riazi-Sahhaf, Eq. (2.42); and Table 2.6.

50

o API Data
40 \~ - - - Winn
* ~ - - - - R-D
- - - - API
Ext. R-D
- - - - L - K

- - - P-F
30
- R-S

~ 20
(,9

10

r ~ i I i t L i ~ I I I I I I I I I ~ I i I I I I I i

5 10 15 20 25 30 35 40 45 50
Carbon Number

FIG. 2.18~Comparison of various methods for estimation of critical pressure of n-

alkylcyclopentanes. API Data: API-TDB [2]; Winn: Eq. (2.95); R-D: Riazi-Daubert, Eq. (2.64);
APh Eq. (2.56); Ext. R-D: Extended Riazi-Daubert, Eq. (2.68); L-K: Lee-Kesler, Eq. (2.70);
P-F: Plan-Firoozabadi, Eq. (2.43); and Table 2.8; R-S: Riazi-Sahhaf, Eq. (2.42); and Table 2.6.
 2. CHARACTERIZATION AND P R O P E R TIE S OF PURE H Y D R O C A R B O N S 79

3000

o API Data /
S
/ .
.
2500 TWO /'i /
.... Ext.R-D /. "~

2000 . . . . . R-S ,//~d- ~ / - ' / ~

- - - --R-D ~Z../ /

1500

= 1000

500

i i I I r i i I i i i i i i i l i i

0 10 20 30 40 50
Carbon Number
FIG. 2.19--Comparison of various methods for estimation of critical volume of
n-alkylbenzenes. API Data: API-TDB [2]; Twu: Eqs. (2.83)-(2.85); APh Eq. (2.101);
R-S: Riazi-Sahhaf, Eq. (2.42) and Table 2.6; R-D: Riazi-Daubert, Eq. (2.98); H-Y:
HalI-Yarborough, Eq. (2.100).

E v a l u a t i o n of these m e t h o d s for critical p r o p e r t i e s of hy- Riazi a n d D a u b e r t correlations p r e s e n t e d b y Eq. (2.63) a n d

d r o c a r b o n s heavier t h a n C20 was n o t possible due to the lack
of confirmed e x p e r i m e n t a l data. Application of these m e t h o d s
for critical p r o p e r t i e s of p e t r o l e u m fractions a n d reservoir
fluids is b a s e d on the a c c u r a c y of p r e d i c t e d physical p r o p -
erty. These evaluations are discussed in C h a p t e r 3, where the
m e t h o d of p s e u d o c o m p o n e n t is i n t r o d u c e d for the estima-
tion of p r o p e r t i e s of p e t r o l e u m fractions. Generally, a m o r e
a c c u r a t e correlation for p r o p e r t i e s of p u r e h y d r o c a r b o n s does
n o t necessarily give better p r e d i c t i o n for p e t r o l e u m fractions
especially those c o n t a i n i n g heavy c o m p o u n d s .
E v a l u a t i o n of m e t h o d s of e s t i m a t i o n of critical p r o p e r t i e s
for p e t r o l e u m fractions is a difficult t a s k as the results d e p e n d
on the type of p e t r o l e u m fraction u s e d for the evaluation. The
(2.64) or the API m e t h o d s p r e s e n t e d by Eqs. (2.65) a n d (2.66)
were developed b a s e d on critical p r o p e r t y d a t a from C5 to C18;
therefore, their a p p l i c a t i o n to p e t r o l e u m fractions c o n t a i n i n g
very heavy c o m p o u n d s w o u l d be less accurate. The K e s l e r -
Lee a n d the Twu m e t h o d were originally developed b a s e d on
s o m e calculated d a t a for critical p r o p e r t i e s of heavy hydrocar-
bons a n d the consistency of To a n d Pc were observed at Pc = 1
a t m at w h i c h T6 was set equal to To. Twu u s e d s o m e values of
Tc a n d Pc b a c k - c a l c u l a t e d from v a p o r p r e s s u r e d a t a for hydro-
c a r b o n s heavier t h a n C20 to extend a p p l i c a t i o n of his corre-
lations to heavy h y d r o c a r b o n s . Therefore, it is expected t h a t
for heavy fractions or reservoir fluids c o n t a i n i n g heavy com-
p o u n d s these m e t h o d s p e r f o r m b e t t e r t h a n Eqs. (2.63)-(2.66)

TABLE 2.15--Evaluation of various methods for prediction of critical temperature and pressure
of pure hydrocarbons from C5 to C20.
Abs Dev%**
T~ Pc
Method Equation(s) AD% MAD% AD% MAD%
API (2.65)-(2.66) 0.5 2.2 2.7 13.2
Twu (2.73)-(2.88) 0.6 2.4 3.9 16.5
Kesler-Lee (2.69)-(2.70) 0.7 3.2 4 12.4
Cavett (2.71 )-(2.73) 3.0 5.9 5.5 31.2
Winn (Sim-Daubert) (2.94)-(2.95) 1.0 3.8 4.5 22.8
Riazi-Daubert (2.63)-(2.64) 1.1 8.6 3.1 9.3
Lin & Chao Reference [72] 1.0 3.8 4.5 22.8
aData on Tc and Pc of 138 hydrocarbons from different families reported in API-TDBwere used for the evalua-
tion process [29].
bDefinedby Eqs. (2.134) and (2.135).
80 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
for Tc and Pc as observed by some researchers [51, 85]. How-
ever, Eq. (2.42) and subsequently derived Eqs. (2.67) and
(2.68) have the internal consistency and can be used from
C20
C5 to C50 although they are developed for hydrocarbons from
to C50
The 1980 Riazi-Daubert correlations for Tc and Pc were
generally used and recommended by m a n y researchers for
light fractions (M < 300, carbon number < C22). Yu et al. [84]
used 12 different correlations to characterize the C7+ plus
fraction of several samples of heavy reservoir fluids and bi-
tumens. Based on the results presented on gas-phase compo-
sition, GOR, and saturation pressure, Eqs. (2.63) and (2.64)
showed better or equivalent predictions to other methods.
Whitson [53] made a good analysis of correlations for the crit-
ical properties and their effects on characterization of reser-
voir fluids and suggested the use of Eqs. (2.63) and (2.64) for
petroleum cuts up to C25. But later [51] based on his observa-
tion for phase behavior prediction of heavy reservoir fluids,
he recommended the use of Kesler-Lee or Twu for estimation
of Tc and Pc of such fluids, while for estimation of critical
volume he uses Eq. (2.98). Soreide [52] in an extensive eval-
uation of various correlations for the estimation of critical
properties recommends use of the API-TDB [2] method for
estimation Tc and Pc (Eqs. (2.65) and (2.66)) but he recom-
mends Twu method for the critical volume. His recommenda-
tions are based on phase behavior calculations for 68 samples
of North Sea reservoir fluids. In a recently published Hand-
book o f Reservoir Engineering [48], and calculations made on
phase behavior of reservoir fluids [86], Eqs. (2.65), (2.66) have
been selected for the estimation of critical properties of unde-
fined petroleum fractions. Another possibility to reduce the
3.00
API Data: n-Alkanes
-- -- Predicted: n-Alkanes
9 API D auc n-Alkylcyclopentanes:
2.50
.... Predicted: n-Alkylcyclopentanes
9 API Data: n-Alkylbenz enes
Preclicted~n-~dkylbeazenes
2.03
1.50
~9
<
1.03
(150
I t I I I r
(103 r
0 10 ~!0
Carbon Number
FIG. 2.20---Prediction of acentric factor of pure h y d r o c a r b o n s from Eq. (2.42).
error associated with critical properties of heavy fractions is
to back-calculate the critical properties of the heaviest end of
the reservoir fluid from an EOS based on a measured physical
property such as density or saturation pressure [51, 52, 70].
Firoozabadi et al. [63, 64] have studied extensively the wax
and asphaltene precipitation in reservoir fluids. They ana-
lyzed various methods of calculating critical properties of
heavy petroleum fractions and used Eq. (2.42) for the critical
properties and acentric factor of paraffins, naphthenes, and
aromatics, but they used Eq. (2.43) for the critical pressure of
various hydrocarbon groups with M > 300. Their evaluation
was based on the calculation of the cloud point of different
oils. It is believed that fractions with molecular weight greater
than 800 (NC57) mainly contain aromatic hydrocarbons [63]
and therefore Eq. (2.42) with constants given in Table 2.6 for
aromatics is an appropriate correlation to estimate the prop-
erties of such fractions.
More recently Jianzhong et al. [87] reviewed and evalu-
ated various methods of estimation of critical properties of
petroleum and coal liquid fractions. Their work followed
the work of Voulgaris et al. [88], who recommended use of
Eq. (2.38) for estimation of critical properties for the purpose
of prediction of physical properties of petroleum fractions
and coal liquids. They correctly concluded that complexity
of correlations does not necessarily increase their accuracy.
They evaluated Lee-Kesler, Riazi-Daubert, and Twu methods
with more than 318 compounds (> C5) including those found
in coal liquids with boiling point up to 418~ (785 ~F) and spe-
cific gravity up to 1.175 [87]. They suggested that Eq. (2.38)
is the most suitable and accurate relation especially when
the coefficients are modified. Based on their database, they
/
/
~j..//
J
: f5
I I
30 40 5o 03
 2. C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P U R E H Y D R O C A R B O N S
obtained the coefficients for Tc, Pc, and Vo in Eq. (2.38) with
use of d20 (liquid density at 20~ and 1 atm in g/cm 3) instead
of SG (Tc, Pc, Vc = aTbbd~0). They reported the coefficients as
[87] T j K (a = 18.2394, b = 0.595251, c = 0.347420), Pc/bar
(a = 2.95152, b = -2.2082, c = 2.22086), and Vc/cma/mol (a =
8.22382 x 10-5, b = 2.51217, c = -1.62214). Equation (2.38)
with these coefficients have not been extensively tested
against data on properties of petroleum fractions as yet but
for more than 300 pure hydrocarbons gives average errors of
0.7, 3.8, and 2.9% for To Pc, and Vc, respectively [87].
2.9.4 E v a l u a t i o n o f M e t h o d s of Estimation
of Acentric Factor and Other Properties
For the calculation of the acentric factor of pure hydrocar-
bons Eq. (2.42) is quite accurate and will be used in Chapter 3
for the pseudocomponent method. Firoozabadi suggests that
for aromatics with M > 800, co = 2. Generally there are three
methods for the estimation of the acentric factor of undefined
petroleum fractions. Perhaps the most accurate method is to
estimate the acentric factor through its definition, Eq. (2.10),
and vapor pressure estimated from a reliable method [86].
This method will be further discussed in Chapter 7 along with
methods of calculation of vapor pressure. For pure hydrocar-
bons the Lee-Kesler method is more accurate than the Edmis-
ter method [36]. The Korsten method for estimating acentric
factor is new and has not yet been evaluated extensively. For
three different homologous hydrocarbon families from C6 to
C50, values of acentric factor calculated from Eq. (2.42) are
compared with values reported in the API-TDB [2] and they
are shown in Fig. 2.20. Prediction of acentric factors from dif-
ferent methods for n-allcylcyclopentanes and n-alkylbenzenes
2.0
o API Data
Riazi- 5ahhaf
.......Lee-Kesler
1.5 Kesler-Lee ~,,"..
.... ~'= .~7:-:'~.""
- -
O
O
1.0 @~176~
< the most appropriate method. The accuracy of a method to
0.5
- 0.0
0 10 20 30
Carbon Number
FIG. 2.21--Prediction of acentric factor of n-alkylcyclo-
pentanes from various methods. API Data: API-TDB [2]; R-S:
Riazi-Sahhaf, Eq. (2.42) and Table 2.6; L-K: Lee-Kesler,
Eq. (2.105); K-L: Kesler-Lee, Eq. (2.107); Edmister: Eq. (2.108);
Korsten: Eq. (2.109).
81
3.000
o DIPPR Data
Riazi-Sahhaf
2.500 P ~ et al. /
9- Korstea /
2.O00 /
m ~ Lee-Kesler /
/.:.--:...---
< 1.000
0.500
0.000
0 10 20 30 40 50
Carbon Number
FIG. 2.22--Prediction of acentric factor of n-alkylbenzenes
from various methods. DIPPR Data: DIPPR [20]; Riazi-Sahhaf:
Eq. (2.42) and Table 2.6; Pan et al.: Ref. [63, 64], Eq. (2.44);
Korsten: Eq. (2.109); Lee-Kesler: Eq. (2.105); Kesler-Lee:
Eq. (2.107).
are presented in Figs. 2.21 and 2.22, respectively. The Riazi-
Sahhaf method refers to Eq. (2.42) and coefficients given
in Table 2.6 for different hydrocarbon families. In Fig. 2.22
the Pan et al. [63, 64] method refers to Eq. (2.44), which
has been recommended for n-alkylbenzenes (aromatics). The
Lee-Kesler method, Eq. (2.105), has been generally used for
the estimation of accentric factor of undefined petroleum
fractions [27]. The Kesler-Lee method refers to Eq. (2.107),
which was recommended by Kesler-Lee [12] for estimation
of the acentric factor of hydrocarbons with Tbr > 0.8, which
/ is nearly equivalent to hydrocarbons with molecular weights
/ / greater than 300. However, our experience shows that this
~. -~ equation is accurate for pure compounds when true critical
temperatures are used and high errors can occur when the
predicted critical temperature is used in the equation. For
heavy hydrocarbons and petroleum fractions (M > 300) with
estimated critical properties, either the method of pseudo-
component discussed in Chapter 3 or the Lee-Kesler may be
estimate acentric factor also depends on the values of Tc and
Pc used to calculate co as was shown in Example 2.7. Usually
the Cavett correlations for Tc and Pc are used together with
the Edmister method. Evaluation of these methods for the
prediction of properties of undefined petroleum fractions is
discussed in Chapter 3.
40 50 The accuracy of correlations presented for estimation of
other properties such as density, refractive index, boiling
point, and CH has been discussed in the previous section
where these methods are presented. Prediction of the refrac-
tive index for pure hydrocarbons is shown in Fig. 2.9. Predic-
tion of viscosity at 38~ (100~ I ) 3 8 , through Eq. (2.128) for
pure hydrocarbons from three hydrocarbon groups is shown
in Fig. 2.23. Further assessment of accuracy of these methods
82 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
30
Z~ API Data: n-Alkanes
Predicted: n-Alkanes
25 o API Data: n-alkyleyclopentanes
..... Predict ed: n-alkylcyclop entanes
9 API Data: n-alkylcbenzenes
20
-- -- Predieted:n-allq,lcbenzenes
15
o
10
//o
. . . . ~ .
0 5 10 15
Carbon Number
FIG. 2.23--Prefliction of kinematic viscosity of pure hydrocarbons at 37.8~
from Eq. (2.128).
is discussed in the following chapters where properties of
petroleum fractions are calculated.
2.10 CONCLUSIONS AND
RECOMMENDATIONS
In this chapter methods of characterization of pure hydro-
carbons have been presented. These methods will be used
in Chapters 3 and 4 for the characterization of petroleum
fractions and crude oils, respectively. This chapter is an im-
portant chapter in this book as the method selected for the
characterization of hydrocarbons affects the accuracy of es-
timation of every physical property throughout the book. In
this chapter the basic characterization parameters have been
introduced and based on the theory of intermolecular forces,
a generalized correlation for the characterization of hydro-
carbon systems has been developed. It is shown that funda-
mentally developed correlations are simpler with a wider field
of application and accuracy. For light fractions (M < 300),
generally two-parameter correlations are sufficient for prac-
tical calculations, while for heavier hydrocarbons or nonhy-
drocarbons the use of a third parameter is needed. The two
characterization parameters should represent the energy and
size characteristics of molecules. Characterization parame-
ters such as Tb, M, and v3s000) may be used as energy, param-
eters while SG, I, and CH could be used as size parameters.
Boiling point and specific gravity are the most easily measur-
able and appropriate characterization parameters followed
by molecular weight and refractive index. Viscosity and CH
i
/
/
/
/
/
/ /
/
1/.,
/
/../
/ ,' /
/.'o
i . . . . i . . . . J . . . .
20 25 30 35
parameters may be used as the last option for prediction of
properties of hydrocarbons. Various methods of estimation
of these parameters as well as critical properties and acen-
tric factor used in corresponding state correlations and a de-
tailed review of their application for different purposes and
recommendations made in the literature are presented. Basic
properties of more than 100 selected compounds are given in
Tables 2.1 and 2.2 and will be used frequently throughout the
book.
The most important information presented in this chap-
ter is the methods of estimation of molecular weight, critical
constants, and acentric factor for pure hydrocarbons. These
methods are also recommended to estimate properties of nar-
row boiling range petroleum fractions as discussed in Chapter
3. A summary of evaluations made by various researchers was
reviewed in Section 2.9. Based on these evaluations it is clear
that theoretically based correlations such as Eq. (2.38) or its
modified version Eq. (2.40), while simpler than other empiri-
cally developed correlations, have a wide range of application
with reasonable accuracy. Based on these evaluations a list
of recommended methods for different properties of various
types of hydrocarbons and narrow boiling range fractions is
given in Table 2.16. Estimation of wide boiling range frac-
tions is discussed in the next chapter. The choice for meth-
ods of calculation of properties not presented in Table 2.16
is generally narrow and comments have been made where
the methods are introduced in each section. The information
presented in this chapter should help practical engineers to
develop new correlations or to select an appropriate charac-
terization scheme when using a process simulator.
 2. CHARACTERIZATION AND PROPERTIES OF PURE HYDROCARBONS 83

TABLE 2.16--Recommended methods for the prediction of the basic properties of pure hydrocarbons
and narrow boiling range petroleum fractionsa.
Property Range of M Method Equation
M 70-700 API [2] 2.51
70-300 Riazi-Dabubert [28] 2.50
200-800 API [2] 2.52(b)
70-700 Twu [30] 2.89-2.92(c)
Tc 70-300 API [2] 2.65
70-700 Lee-Kesler [ 12] 2.69
70-700 Extended API [65] 2.67
70-800 Riazi-Sahhaf [31 ] 2.42 d
<70 e Riazi et al. [37] 2.47 e
Pc 70-300 API [2] 2.66
70-700 Lee-Kesler [12] or Twu [30] 2.70
70-700 Extended API [65] 2.68
70-300 Riazi-Sahhaf [31 ] 2.42 d
300-800 Pan-Firoozabadi-Fotland [63] 2.43 d
<70 e Riazi et al. [37] 2.47 e
Vc 70-350 Riazi-Daubert [28] 2.98
300-700 Extended R-D [65] 2.99
70-700 Riazi-Sahhaf [31] 2.42 d
<70 e Riazi et al. [37] 2.47 e
Zc 70-700 By definition of Zc 2.8
w 70-300 Lee-Kesler [27] 2.105
300-700 Korsten [77] 2.109
70-700 Riazi-Sahhaf [31 ] 2.42 f
300-700 Pan-Firoozabadi-Fofland [63] 2.44g
Tb 70-300 Riazi-Daubert [29] 2.56
300-700 Extended R-D [65] 2.57
70-700 Riazi-Sahhaf [31 ] 2.42 d
SG All range Denis et al. [8] 2.112
70-300 Riazi-Daubert [29] 2.59
70-700 Extended R-D [65] 2.60
200-800 API [2] 2.61 d
I 70-300 Riazi-Daubert [29] 2.116
300-700 Extended R-D [65] 2.117
70-700 Riazi-Sahhaf [31 ] 2.42 d
d All range Denis et al. [8] 2.111
70-350 Riazi-Daubert [28] 2.113
TM 70-700 Pan-Firoozabadi-Fotland [63] 2.126 h
Riazi-Sahhaf [31] 2.124 and 2.125 i
Methods recommended for pure homologous hydrocarbons (designated by c-i) are also recommended for the pseu-
docomponent method discussed in Chapter 3 for petroleum fractions. The 300 boundaryis approximate and methods
recommended for the range of 70-300 may be used safely up to molecular weight of 350 and similarly methods rec-
ommended for the range 300-700 may be used for molecular weight as low as 250.
a For narrow boiling range fractions a midpoint distillation temperature can be used as Tb.
bOnly when Tb is not available.
CRecommended for pure hydrocarbons from all types.
'/Recommended for pure homologous hydrocarbon groups.
eFor compounds and fractions with molecular weight less than 70 and those containing nonhydrocarbon compounds
(H2S,CO2, N2, etc.) Eq. (2.47) is recommended.
fEquation (2.42) is applicable to acentric factor of n-alkylbenzenes up to molecular weight of 300.
gEquation (2.44) is applicable to acentric factor of aromatics for 300 < M < 800 and for M > 800, w - 2 should be
used.
hFor pure hydrocarbons from n-alkanes family.
i For pure hydrocarbons from n-alkylcylopentanes (naphthenes) and n-alkylbenzenes (aromatics) families.

2.11 P R O B L E M S 2.2. F o r h e a v y a n d c o m p l e x h y d r o c a r b o n s o r p e t r o l e u m frac-

tions, b a s i c p r o p e r t i e s c a n b e b e s t d e t e r m i n e d f r o m
2.1. F o r light h y d r o c a r b o n s a n d n a r r o w b o i l i n g r a n g e frac- three parameters. Determine the best three parameters
t i o n s u s u a l l y a few m e a s u r e d p a r a m e t e r s a r e avail- for e a c h of t h e f o l l o w i n g cases:
able. F o r e a c h o n e of t h e f o l l o w i n g cases d e t e r m i n e a. Tb, M, SG, I)38(100)
t h e b e s t t w o p a r a m e t e r s f r o m t h e set of a v a i l a b l e b. C H , 1)38(100), n20, 1)99(210), API G r a v i t y
d a t a t h a t are s u i t a b l e to b e u s e d for p r o p e r t y predi- c. CH, n20, SG, M, 1)99(210)
ctions: d. Tb, M, n20 , CH, Kw
a. Tb, M, SG 2.3. You w i s h to d e v e l o p a p r e d i c t i v e c o r r e l a t i o n f o r p r e d i c -
b. C H , 1)38(100), n20 t i o n of m o l a r v o l u m e , Vr, in t e r m s of v38(100), SG, a n d
c. CH, n20, S G t e m p e r a t u r e T. H o w do y o u p r o p o s e a s i m p l e r e l a t i o n
d. Tb, M, n20, C H w i t h t e m p e r a t u r e d e p e n d e n t p a r a m e t e r s for e s t i m a t i o n
e. 1338(100), Tb, C H , M of m o l a r v o l u m e ?
84 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
2.4. A tank contains pure h y d r o c a r b o n liquid from the
n-paraffin group. Determine the lightest h y d r o c a r b o n
from the n-alkane family that can exist in an open ves-
sel at the environment of 38~ (100~ and 1 atm with-
out danger of flammability in the vapor phase near the
vessel.
2.5. Develop three relations for estimation of CH weight
ratio of n-paraffins, n-alkylcyclopentanes, and n-alkyl-
benzene in terms of their respective molecular weight.
For each group calculate CHo~ (and HCo~). Show graph-
ical presentation of the predicted values versus actual
values of CH for the three families on a single graph.
2.6. Predict the refractive index of n-paraffins, n-alkyl-
cyclopentanes, and n-alkylbenzene versus carbon num-
ber from C6 to Cs0 using Eq. (2.46a) and c o m p a r e graph-
ically with values from Eq. (2.42). In using Eq. (2.46a)
it is necessary to obtain M from Nc in each family, and
then from Eq. (2.42) Tb and SG m a y be estimated for
each carbon n u m b e r in each family.
2.7. A pure hydrocarbon has molecular weight of 338.6 and
specific gravity of 0.8028. Using appropriate methods
calculate
a. boiling point, Tb.
b. refractivity intercept, Ri.
c. kinematic viscosity at 38 and 99~
d. VGC f r o m three different methods.
2.8. For n-butylcyclohexane, critical properties and molecu-
lar weight are give in Table 2.1. Use Tb and SG as the
input partameters and calculate
a. M, To, Pc, de, and Zc from the API-TDB-87 methods.
b. M, To, Pc, dc, and Zc the Lee-Kesler correlations.
c. M, To, Pc, de, and Zc from the Riazi-Daubert correla-
tions (Eq. 2.38).
d. M, Tc, Pc, de, and Zc from the Twu correlations.
e. Compare values from each m e t h o d with actual values
and tabulate the %D.
2.9. Use calculated values of Tc and Pc in Problem 2.8 to cal-
culate acentric factor from the Lee-Kesler and Korsten
correlations for each part, then obtain the errors (%D)
for each method.
2.10. Estimate the acentric factor of isooctane from Lee-
Kesler, Edmister, and Korsten correlations using input
data from Table 2.1. Calculate the %D for each method.
2.11. Estimate the kinematic viscosity of n-heptane at 38
and 99~ and c o m p a r e with the experimental values
reported by the API-TDB [2]. Also estimate viscosity
of n-heptane at 50~ from Eq. (2.130) and the ASTM
viscosity-temperature chart.
2.12. For n-alkylcylopentanes from C5 to C10, estimate d20
from SG using the rule of thumbs and a more accurate
method. Compare the results with actual values from
Table 2.1. For these c o m p o u n d s also estimate refractive
index at 25~ using M as the only input data available.
Use both methods for the effect of temperature on re-
fractive index as discussed in Section 2.6.2 and c o m p a r e
your results with the values reported by the API-TDB [2].
REFERENCES
[1] Coplen, T. B., "Atomic Weights of the Elements 1999," Pure and
Applied Chemistry, Vol. 73, No. 4, 2001, pp. 667-683.
[2] Daubert, T. E. and Danner, R. E, Eds., API Technical Data
Book--Petroleum Refining, 6th ed., American Petroleum
Institute (API), Washington, DC, 1997.
[3] Nikitin, E. D., Pavlov, E A., and Popov, A. E, '%'apor-Liquid
Critical Temperatures and Pressures of Normal Alkanes from
19 to 36 Carbon Atoms, Naphthalene and m-Terphenyl
Determined by the Pulse-Heating Technique," Fluid Phase
Equilibria, Vol. 141, 1997, pp. 155-164.
[4] Reid, R. C., Pransnitz, J. M., and Poling, B. E., Properties of
Gases and Liquids, 4th ed., Mc-Graw Hill, New York, 1987.
Poling, B. E., Pransnitz, J. M., O'Connell, J. E, Properties of
Gases and Liquids, 5th ed., Mc-Graw Hill, New York, 2001.
[5] Pitzer, K. S., "The Volumetric and Thermodynamic Properties
of Fluids, I: Theoretical Basis and Virial Coefficients," Journal
of American Chemical Society, Vol. 77, 1955, pp. 3427-3433.
[6] Pitzer, K. S., Lippmann, D. Z., Curl, Jr., R. E, Huggins, C. M.,
and Petersen, D. E., "The Volumetric and Thermodynamic
Properties of Fluids, II: Compressibility Factor, Vapor Pressure,
and Entropy of Vaporization," Journal of American Chemical
Society, Vol. 77, 1955, pp. 3433-3440.
[7] Denis, J., Briant, J., and Hipeaux, J. C., Lubricant Properties
Analysis and Testing, Translated to English by G. Dobson,
Editions Technip, Paris, 1997.
18] Wauquier, J.-P., Petroleum Refining, Vol. 1: Crude Oil, Petroleum
Products, Process Flowsheets, Translated from French by David
H. Smith, Editions Technip, Paris, 1995.
[9] Watson, K. M., Nelson, E. E, and Murphy, G. B.,
"Characterization of Petroleum Fractions," Industrial and
Engineering Chemistry, Vol. 27, 1935, pp. 1460-1464.
[10] Daubert, T. E., "Property Predictions," Hydrocarbon Processing,
Vol. 59, No. 3, 1980, pp. 107-112.
[11] Speight, J. G., The Chemistry and Technology of Petroleum, 3rd
ed., Marcel Dekker, New York, 1998.
[12] Kesler, M. G. and Lee, B. I., "Improve Prediction of Enthalpy of
Fractions," Hydrocarbon Processing, Vol. 55, 1976,
pp. 153-158.
[13] Kurtz, Jr., S. S. and Ward, A. L., "The Refractivity Intercept and
the Specific Refraction Equation of Newton, I: Development of
the Refractivity Intercept and Composition with Specific
Refraction Equations," Journal of Franklin Institute, Vol. 222,
1936, pp. 563-592.
[14] Hill, J. B. and Coats, H. B., "The Viscosity Gravity Cosntant of
Petroleum Lubricating Oils," Industrial and Engineering
Chemistry, Vol. 20, 1928, pp. 641-4544.
[15] Riazi, M. R. and Daubert, T. E., "Prediction of the Composition
of Petroleum Fractions," Industrial and Engineering Chemistry,
Process Design and Development, Vol. 19, No. 2, 1980,
pp. 289-294.
[16] Kurtz, S. S., King, R. W., Stout, W. J., Partikian, D. G., and
Skrabek, E. A., "Relationsship Between Carbon-Type
Composition, Viscosity Gravity Constant, and Refractivity
Intercept of Viscous Fractions of Petroleum," Analytical
Chemistry, Vol. 28, 1956, pp. 1928-1936.
[17] Altgelt, K. H. and Boduszynski, M. M., Composition and
Analysis of Heavy Petroleum Fractions, Marcel Dekker, New
York, 1994.
[I 8] API Research Project 42: Properties of Hydrocarbons of High
Molecular Weight, American Petroleum Institute, New York,
1966.
[19] Daubert, T. E., Danner, R. E, Sibul, H. M., and Stebbins, C. C.,
Physical and Thermodynamic Properties of Pure Compounds:
Data Compilation, DIPPR-AIChE, Taylor & Francis, Bristol, PA,
1994 (extant) (www.aiche.org/dippr), Updated reference:
Rowley, R. L., Wilding, W. V., Oscarson, J. L., Zundel, N. A.,
Marshall, T. L., Daubert, T. E., and Danner, R. E, DIPPR Data
Compilation of Pure Compound Properties, Design Institute for
Physical Properties (DIPPR), Taylor & Francis, New York, 2002
(http://dippr.byu.edu).
 2. C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P U R E H Y D R O C A R B O N S
[20] AIChE DIPPR| Design Institute for Physical
Property Data (DIPPR), EPCON International, Houston, 1996.
[21] Hall, K. R., Ed., TRC Thermodynamic Tables--Hydrocarbons,
Thermodynamic Research Center, The Texas A&M University
System, College Station, TX, 1993. Updated reference: TRC
Thermodynamic Tables--Hydrocarbons, Thermodynamic
Research Center, National Institute of Standards and
Technology, Boulder, CO, 2001. (http://www.trc.nist.gov/).
[22] API Research Project 44: Selected Values of Properties of
Hydrocarbons and Related Compounds, Tables of Physcial and
Thermodynamic Properties of Hydrocarbons, A&M Press,
College Station, TX, 1978 (extant).
[23] Dortmund Data Bank (DDB), Laboratory for Thermophysical
Properties, LTP GmbH, Institute at the University of Oldenburg
(www.ddbst.de).
[24] Watson, K. M. and Nelson, E. E, "Improved Methods for
Approximating Critical and Thermal Properties of Petroleum
Fractions," Industrial and Engineering Chemistry, Vol. 25, 1933,
pp. 880-887.
[25] Winn, E W., "Physical Properties by Nomogram," Petroleum
Refiners, Vol. 36, No. 21, 1957, pp. 157-159.
[26] Cavett, R. H., "Physical Data for Distillation Calculations,
Vapor-Liquid Equilibria," Proceeding of 27th API Meeting, API
Division of Refining, Vol. 42, No. 3, 1962, pp. 351-366.
[27] Lee, B. I. and Kesler, M. G., "A Generalized Thermodynamic
Correlation Based on Three- Parameter Corresponding States,"
American Institute of Chemical Engineers Journal, Vot. 21, 1975,
pp. 510-527.
[28] Riazi, M. R. and Daubert, T. E., "Simplify Property Predictions,"
Hydrocarbon Processing, Vol. 59, No. 3, 1980, pp. 115-116.
[29] Riazi, M. R. and Daubert, T. E., "Characterization Parameters
for Petroleum Fractions," Industrial and Engineering Chemistry
Research, Vol. 26, 1987, pp. 755-759. (Corrections, p. 1268.)
[30] Twu, C. H., "An Internally Consistent Correlation for Predicting
the Critical Properties and Molecular Weights of Petroleum and
Coal-Tar Liquids," Fluid Phase Equilbria, Vol. 16, 1984,
pp. 137-150.
[31] Riazi, M. R. and A1-Sahhaf, T., "Physical Properties of n-Alkanes
and n-Alkyl Hydrocarbons: Application to Petroleum Mixtures,"
Industrial and Engineering Chemistry Research, Vol. 34, 1995,
pp. 4145-4148.
[32] Korsten, H., "Characterization of Hydrocarbon Systems by
DBE Concept," American Institute of Chemical Engineers
Journal, Vol. 43, No. 6, 1997, pp. 1559-1568.
[33] Korsten, H., "Critical Properties of Hydrocarbon Systems,"
Chemical Engineering and Technology, Vol. 21, No. 3, 1998,
pp. 229-244.
[34] Tsonopoulos, C., Heidman, J. L., and Hwang, S.-C.,
Thermodynamic and Transport Properties of Coal Liquids, An
Exxon Monograph, Wiley, New York, 1986.
[35] Riazi, M. R. and Roomi, Y., "Use of the Refractive Index in the
Estimation of Thermophysical Properties of Hydrocarbons and
Their Mixtures," Industrial and Engineering Chemistry
Research, Vol. 40, No. 8, 2001, pp. 1975-1984.
[36] Riazi, M. R., Prediction of Thermophysical Properties of
Petroleum Fractions, Doctoral Dissertation, Department of
Chemical Engineering, Pennsylvania State University,
University Park, PA, 1979.
[37] Riazi, M. R., Al-Sahhaf, T. A., and A1-Shammari, M. A.,
"A Generalized Method for Estimation of Critical Constants,"
Fluid Phase Equilibria, Vol. 147, 1998, pp. 1-6.
[38] Hirschfelder, O. J., Curtiss, C. E, and Bird, R. B., Molecular
Theory of Gases and Liquids, Wiley, New York, 1964.
[39] Prausnitz, J. M., Lichtenthaler, R. N., and de Azevedo, E. G.,
Molecular Thermodynamics of Fluid-Phase Equilibria,
Prentice-Hall, New Jersey, 1986.
[40] Hill, T. L., An Introduction to Statistical Thermodynmaics,
Addison Wesley, Boston, MA, 1960.
85
[41] Smith, J. M., Van Ness, H. C., and Abbott, M. M., Introduction
to Chemical Engineering Thermodynamics, 5th ed.,
McGraw-Hill, New York, 1996.
[42] Huang, E K., Characterization and Thermodynamic Correlations
for Undefined Hydrocarbon Mixtures, Ph.D. Dissertation,
Pennsylvania State University, University Park, PA, 1977.
[43] Ambrose, D., "Correlation and Estimation of Vapor-Liquid
Critical Properties, I: Critical Temperatures of Organic
Compounds," National Physical Laboratory, Teddington,
Middlesex, UK, TW11 0LW, NPL Report, Vol. 92, 1978, NPL
Report 98, 1979.
[44] Nokay, R., "Estimate Petrochemical Properties," Chemical
Engineering, Vol. 6, No. 4, 1959, pp. 147-148.
[45] Spencer, C. E and Daubert, T. E., "A Critical Evaluation of
Methods for the Prediction of Critical Propertis of
Hydrocarbons," American Institute of Chemical Engineers
Journal, Vol. 19, No. 3, 1973, pp. 482486.
[46] Edmister, W. C. and Lee, 13. I., Applied Hydrocarbon
Thermodynamics, 2nd ed., Gulf Publishing, Houston, TX,
1985.
[47] Stange, E. and Johannesen, S. O., "HYPO*S, A Program for
Heavy End Characterization," An Internal Program from the
Norsk Hydro Oil and Gas Group (Norway), Paper presented at
the 65th Annual Convection, Gas Processors Association,
San Antonio, TX, March 10, 1986.
[48] Ahmed, T., Reservoir Engineering Handbook, Gulf Publishing,
Houston, TX, 2000.
[49] Ahmed, T., Hydrocarbon Phase Behavior, Gulf Publishing,
Houston, TX, 1989.
[50] "API Tech Database Software," EPCON International, Houston,
TX, 2000 (www.epcon.com).
[51] Whitson, C. H. and Brule, M. R., Phase Behavior, Monograph
Vol. 20, Society of Petroleum Engineers, Richardson, TX,
2000.
[52] Soreide, I., Improved Phase Behavior Predictions of Petroleum
Reservoir Fluids from a Cubic Equation of State, Doctor of
Engineering Dissertation, Norwegian Institute of Technology,
Trondheim, Norway, 1989.
[53] Whitson, C. H., "Effect of C7+ Properties on Equation-of-State
Predictions," Society of Petroleum Engineers Journal, December
1984, pp. 685-696.
[54] Aspen Plus, "Introductory Manual Software Version," Aspen
Technology, Inc., Cambridge, MA, December 1986.
[55] PRO/II, "Keyword Manual," Simulation Sciences Inc.,
Fullerton, CA, October 1992.
[56] HYSYS, "Reference Volume 1, Version 1.1," HYSYS Reference
Manual for Computer Software, HYSYS Conceptual Design,
Hyprotech Ltd., Calgary, Alberta, Canada, 1996.
[57] Kreglewski, A. and Zwolinski, B. J., "A New Relation for
Physical Properties of n-Alkanes and n-Alkyl Compounds,"
Journal of Physical Chemistry, Vol. 65, 1961, pp. 1050-1052.
[58] Tsonopoulos, C., "Critical Constants of n-Alkanes from
Methane to Polyethylene," American Institute of Chemical
Engineers Journal, Vo. 33, No. 12, 1987, pp. 2080-2083.
[59] Gray, Jr., R. D., Heidman, J. L., Springer, R. D., and
Tsonopolous, C., "Characterization and Property Prediction for
Heavy Petroleum and Sysnthetic Liquids," Fluid Phase
Equilibria, Vol. 53, 1989, pp. 355-376.
[60] Teja, A. S., Lee, R. J., Rosenthal, D. J., and Anselme, M.,
"Correlation of the Critical Properties of n-Alkanes and
Alkanols," Fluid Phase Equilibria, Vol. 56, 1990,
pp. 153-169.
[61] Goossens, A. G., "Prediction of Molecular Weight of Petroleum
Fractions," Industrial and Engineering Chemistry Research,
Vol. 35, 1996, pp. 985 988.
[62] Goossens, A. G., "Prediction of the Hydrogen Content of
Petroleum Fractions," Industrial and Engineering Chemistry
Research, Vol. 36, 1997, pp. 2500-2504.
86 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
[63] Pan, H., Firoozabadi, A., and Fotland, E, "Pressure and
Composition Effect on Wax Precipitation: Experimental Data
and Model Results," Society of Petroleum Engineers
Production and Facilities, Vol. 12, No. 4, 1997,
pp. 250-258.
[64] Firoozabadi, A., Thermodynamics of Hydrocarbon Reservoirs,
McGraw Hill, New York, 1999.
[65] Riazi, M. R. and Adwani, H. A., "Some Guidelines for Choosing
a Characterization Method for Petroleum Fractions in Process
Simulators," Chemical Engineering Research and Design,
IchemE, Vol. 83, 2005.
[66] Vetere, A., "Methods to predict the critical constants of organic
compounds," Fluid Phase Equilibria, Vol. 109, 1995,
pp. 17-27.
[67] Riazi, M. R., Daubert, 1". E., "Molecular Weight of Heavy
Fractions from Viscosity," Oil and Gas Journal, Vol. 58, No. 52,
1987, pp. 110-113.
[68] ASTM, Annual Book of Standards, Section Five: Petroleum
Products, Lubricants, and Fossil Fuels (in Five Volumes),
ASTM International, West Conshohocken, PA, 2002.
[69] Hirschler, A. E., "Molecular Mass of Viscous Hydrocarbon Oils:
Correlation of Density with Viscosities," Journal of the Institute
of Petroleum, Vol. 32, No. 1946, 267, pp. 133-161.
[70] Chorn, L. G. and Mansoori, G. A., C7+Characterization, Taylor
& Francis, New York, 1989.
[71] Kesler, M. G., Lee, B. I., and Sandler, S. I., "A Third Parameter
for Use in Generalized Thermodynamic Correlations,"
Industrial and Engineering Chemistry Fundamentals, Vol. 18,
1979, pp. 49-54.
[72] Lin, H. M. and Chao, K. C., "Correlation of Critical Properties
and Acentric Factor of Hydrocarbons and Derivatives,"
American Institute of Chemical Engineers Journal, Vol. 30,
No. 6, 1984, pp. 981-983.
[73] Mobil Oil Company, "Engineering Data Book," New York,
1961.
[74] Sire, W. J. and Daubert, T. E., "Predcition of Vapor-Liquid
Equilibria of Undefined Mixtures," Industrial and Engineering
Chemistry, Process Design and Development, Vol. 19, 1980,
pp. 386-393.
[75] Hall, K. R. and Yarborough, L., "New Simple Correlation for
Predicting Critical Volume," Chemical Engineering, Vol. 78,
No. 25, 1971, p. 76-77.
[76] Edmister, W. C., "Applied Hydrocarbon Thermodynamics, Part
4: Compressibility Factors and Equations of State," Petroleum
Refiner, Vol. 37, 1958, pp. 173-179.
[77] Korsten, H., "Internally Consistent Prediction of Vapor
Pressure and Related Properties," Industrial and Engineering
Chemistry Research, 2000, Vol. 39, pp. 813-820.
[78] Riazi, M. R. and Daubert, 1". E., "Prediction of Molecular Type
Analysis of Petroleum Fractions and Coal Liquids," Industrial
and Engineering Chemistry, Process Design and Development,
Vol. 25, No. 4, 1986, pp. 1009-1015.
[79] Won, K. W., Thermodynamics for Solid Solution-Liquid-Vapor
Equilibria: Wax Formation from Heavy Hydrocarbon
Mixtures," Fluid Phase Equilibria, Vol. 30, 1986, pp. 265-279.
[80] Abbott, M. M., Kaufrnann, 17.G. and Domash, L., "A Correlation
for Predicting Liquid Viscosities of Petroleum Fractions,"
Canadian Journal of Chemical Engineering, Vol. 49, No. 3, 197l,
pp. 379-384.
[81] Twu, C. H., "Internally Consistent Correlation for Predicting
Liquid Viscosities of Petroleum Fractions," Industrial and
Engineering Chemistry, Process Design and Development,
Vol. 24, 1985, 1287-1293.
[82] Riazi, M. R. and Elkamel, A. "Using Neural Network Models to
Estimate Critical Constants", Simulators International,
International Society for Computer Simulation, Vol. 29, No. 3,
1997, pp. 172-177.
[83] Lin, H. M., Kim, H., Gut, T., and Chat, K. C., "Equilibrium
Vaporization of a Coal Liquid from a Kentucky No. 9 Coal,"
Industrial and Engineering Chemistry, Process Design and
Development, Vol. 24, No. 4, 1985, pp. 1049-1055.
[84] Yu, J. M., Wu, R. S., and Batycky, J. E, "A C7+ Characterization
Method for Bitumen Mixture Phase Behavior Prediction,"
C7+Characterization, L. G. Chorn and G. A. Mansoori Eds.,
Taylor & Francis, New York, 1989, pp. 169-196.
[85] Danesh, A., PVT and Phase Behavior of Petroleum Reservoir
Fluids, Elsevier, Amsterdam, 1998.
[86] Manafi, H., Mansoori, G. A. and Ghotbi, S., "Phase Behavior
Prediction of Petroleum Fluids with Minimum
Characterization Data," Journal of Petroleum Science and
Engineering, Vol. 22, 1999, pp. 67-93.
[87] Jianzhong, Zh., Biao, Zh., Suoqi, Zh., Renan, W., and
Guanghua, Y., "Simplified Prediction of Critical Properties of
Nonpolar Compounds, Petroleum, and Coal Liquid Fractions,"
Industrial and Engineering Chemistry Research, Vol 37, No. 5,
1998, pp. 2059-2060.
[88] Voulgaris, M., Stamatakis, S., Magoulas, K., and Tassios, D.,
"Prediction of Physical Properties for nonpolar Compounds,
Petroleum, and Coal Liquid Fractions," Fluid Phase Equilibria,
Vol. 64, 1991, pp. 73-106.
 MNL50-EB/Jan. 2005

Characterization of Petroleum
Fractions

NOMENCLATURE 0 A property of hydrocarbon such as: M, To, Pc, Vc, I, d,

Tb, 9 99

AP Aniline point, ~ (unless specified otherwise) p Density at a given temperature and pressure, g/cm 3
API API gravity defined in Eq. (2.4) a Surface tension, dyn/cm [=mN/m]
A,B, . . . . F Correlation coefficients in various equations o~ Acentric factor defined by Eq. (2.10), dimensionless
a,b..i Correlation coefficients in various equations
CABP Cubic average boiling point, K Superscript
CH Carbon-to-hydrogen weight ratio
~ Properties of n-alkanes from Twu correlations
d Liquid density at 20~ and 1 atm, g/cm 3
Kw Watson (UOP) K factor defined by Eq. (2.13)
I Refractive index parameter defined in Eq. (2.36)
Subscripts
M Molecular weight, g/mol [kg/kmol] A Aromatic
MABP Molal average boiling point, K N Naphthenic
MeABP Mean average boiling point, K P Paraffinic
n Sodium D line refractive index of liquid at 20~ T Value of a property at temperature T
and 1 atm, dimensionless ~ A reference state for T and P
Nc Carbon number (number of carbon atoms in a c~ Value of a property at M --~ c~
hydrocarbon molecule) 20 Value of a property at 20~
Critical pressure, bar 39(100) Value of kinematic viscosity at 37.8~ (100~
Refractivity intercept in Eq. (2.14) 99(210) Value of kinematic viscosity at 98.9~ (210~
SG Specific gravity of liquid substance at 15.5~
(60~ defined by Eq. (2.2), dimensionless Acronyms
SGg Specific gravity of gas substance at 15.5~ (60~
defined by Eq. (2.6), dimensionless API-TDB American Petroleum Institute--Technical Data
SL ASTM D 86 slope between 10 and 90% points, Book
~ (K)/vol% ASTM D ASTM International (test methods by D com-
S% Weight percent of sulfur in a fraction mittee)
Tb Boiling point, K %AD Absolute deviation percentage defined by
Eq. (2.134)
Tc Critical temperature, K
TF Flash point, K %AAD Average absolute deviation percentage defined by
TM Melting (freezing point) point, K Eq. (2.135)
EFV Equilibrium flash vaporization
T~o Temperature on distillation curve at 10% volume
vaporized, K EOS Equation of state
Ts0 Temperature on distillation curve at 50% volume FBP Final boiling point (end point)
vaporized, K GC Gas chromatography
V Molar volume, cma/gmol GPC Gel permeation chromatography
V Saybolt universal viscosity, SUS HPLC High performance liquid chromatography
v~ Critical volume (molar), cm3/gmol KISR Kuwait Institute for Scientific Research
VGC Viscosity gravity constant defined by Eqs. (2.15)- IBP Initial boiling point
(2.18) IR Infrared
VABP Volume average boiling point, K MA Monoaromatic
WABP Weight average boiling point, K MS Mass spectroscometry
PA Poly (di- tri-, and higher) aromatic
PIONA Paraffin, isoparaffin, olefin, naphthene,
Greek Letters and aromatic
RVP Reid vapor pressure
F G a m m a function RS R squared (R2), defined by Eq. (2.136)
# Absolute (dynamic) viscosity, cp [mPa.s]. Also used for SD Simulated distillation
dipole moment TBP True boiling point
v Kinematic viscosity defined by Eq. (2.12), cSt [mm2/s] UV Ultraviolet
87

Copyright 9 2005 by ASTM International www.astm.org

88 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
IN THIS CHAPTERmethods of characterization of petroleum
fractions and products are discussed. Petroleum fractions are
mixtures of hydrocarbon compounds with a limited boiling
point range. Experimental methods on measurement of ba-
sic properties that can be obtained from laboratory testing are
first presented and then methods of prediction of properties
that are not available will be discussed. Two general meth-
ods are presented: one for defined mixtures and another for
undefined mixtures in which the composition is not known
but some bulk properties are available. Petroleum fractions
are also divided into light and heavy as well as narrow and
wide boiling range mixtures in which different characteriza-
tion methods are proposed. In addition to methods of estima-
tion of characterization parameters discussed in Chapter 2
for pure hydrocarbons, predictive methods for some char-
acteristics specifically applicable to petroleum fractions are
presented in this chapter. These characteristic parameters in-
clude distillation curve types and their interconversions, hy-
drocarbon type composition, sulfur content, carbon residue,
octane number, pour, cloud, aniline, and smoke points that af-
fect the quality of a fuel. Standard test methods recommended
by ASTM are given for various properties. Finally, minimum
laboratory data needed to characterize various fractions as
well as analysis of laboratory data and criteria for develop-
ment of a predictive method are discussed at the end of this
chapter. Most of methods presented in this chapter will also
be used in Chapter 4 to characterize crude oils and reservoir
fluids.
3.1 EXPERIMENTAL DATA ON BASIC
PROPERTIES OF PETROLEUM FRACTIONS
In this section characterization parameters that are usually
measured in the laboratory as well as methods of their mea-
surements are discussed. Generally not all of these parame-
ters are reported in a laboratory report, but at least from the
knowledge of some of these properties, all other basic prop-
erties for the fraction can be determined from the methods
presented in this chapter.
3.1.1 B o i l i n g P o i n t a n d D i s t i l l a t i o n C u r v e s
Pure compounds have a single value for the boiling point;
however, for mixtures the temperature at which vaporization
occurs varies from the boiling point of the most volatile com-
ponent to the boiling point of the least volatile component.
Therefore, boiling point of a defined mixture can be repre-
sented by a number of boiling points for the components ex-
isting in the mixture with respect to their composition. For
a petroleum fraction of unknown composition, the boiling
point may be presented by a curve of temperature versus
vol% (or fraction) of mixture vaporized. Different mixtures
have different boiling point curves as shown in Fig. 3.1 for a
gas oil petroleum product [1]. The curves indicate the vapor-
ization temperature after a certain amount of liquid mixture
vaporized based on 100 units of volume, The boiling point of
the lightest component in a petroleum mixture is called ini-
tial boiling point (IBP) and the boiling point of the heaviest
compound is called the final boiling point (FBP). In some ref-
erences the FBP is also called the end point. The difference
330
Gas Oil
320 .... n-Tetradecane
310 n-Hexadecane
m ~ n-Nonadecane
~ 300
2
~ 290
=
280
270
260
250 L t i I
0 20 40 60 80 100
Vol% Vaporized
FIG. 3.1--Distillation curve for a gas oil and three pure
hydrocarbons,
between FBP and IBP is called boiling point range or simply
boiling range. For petroleum fractions derived from a crude
oil, those with wider boiling range contain more compounds
than fractions with narrower boiling range. This is due to
the continuity of hydrocarbon compounds in a fraction. Ob-
viously, in general, for defined mixtures this is not the case.
For a pure component the boiling range is zero and it has
a horizontal distillation curve as shown in Fig. 3.1 for three
n-alkane compounds of C14, C16, and C19. For the gas oil sam-
ple shown in Fig. 3.1 the IBP is 248~ (477~ and the FBP is
328~ (62 I~ Therefore its boiling range is 80~ (144~ and
compounds in the mixture have approximate carbon number
range of C14-C19. Crude oils have boiling ranges of more than
550~ (~1000~ but the FBPs are not accurate. For heavy
residues and crude oils the FBPs may be very large or even
infinite as the heaviest components may never vaporize at
all. Generally, values reported as the IBP and FBP are less
reliable than other points. Ft3P is in fact the maximum tem-
perature during the test and its measurement is especially
difficult and inaccurate. For heavy fractions it is possible
that some heavy compounds do not vaporize and the high-
est temperature measured does not correspond to the boiling
point of heaviest component present in the mixture. If the
temperature is measured until, i.e. 60% vaporized, then the
remaining 40% of the fraction is called residue. The boiling
point curve of petroleum fractions provides an insight into the
composition of feedstocks and products related to petroleum
refining processes. There are several methods of measuring
and reporting boiling points of petroleum fractions that are
described below.
3.1.1.1 A S T M D 86
ASTM D 86 is one of the simplest and oldest methods of mea-
suring and reporting boiling points of petroleum fractions
and is conducted mainly for products such as naphthas,
gasolines, kerosenes, gas oils, fuel oils, and other similar
petroleum products. However, this test cannot be conducted
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS 89
FIG. 3.2--Experimental apparatus for measurement of boiling point of
petroleum fractions by ASTM D 86 method (courtesy of Kuwait Institute
for Scientific Research).
for mixtures containing very light gases or very heavy com-
pounds that cannot be vaporized. The test is conducted at
atmospheric pressure with 100 mL of sample and the result
is shown as a distillation curve with temperatures at 0, 5, 10,
20, 30, 40, 50, 60, 70, 80, 90, 95, and 100% volume vapor-
ized. The final boiling point (at 100%) is the least accurate
value and it is usually less than the true final boiling point. In
many cases only a few temperatures are reported. An exposed
thermometer is used and temperatures are reported without
stem corrections. For heavy products, temperatures are re-
ported at maximum of 90, 70, or even 50% volume vaporized.
This is due to the cracking of heavy hydrocarbons at high tem-
peratures in which vaporization temperatures do not repre-
sent boiling points of the original compounds in the mixture.
The cracking effect is significant at temperatures above 350 ~C
(660~ however, ASTM D 86 temperatures reported above
250~ (480~ should be used with caution. Corrections ap-
plied to consider the effects of cracking are applicable from
250 to 500~ however, these procedures have not been widely
used and generally have not been confirmed. In the new revi-
sions of API-TDB-97 no correction for cracking in ASTM D 86
temperatures has been recommended [2]. An apparatus to
measure distillation of petroleum fractions by ASTM D 86
method is shown in Fig. 3.2.
3.1.1.2 True Boiling Point
ASTM D 86 distillation data do not represent actual boiling
point of components in a petroleum fraction. Process engi-
neers are more interested in actual or true boiling point (TBP)
of cuts in a petroleum mixture. Atmospheric TBP data are
obtained through distillation of a petroleum mixture using a
distillation column with 15-100 theoretical plates at relatively
high reflux ratios (i.e., 1-5 or greater). The high degree of
fractionation in these distillations gives accurate component
distributions for mixtures. The lack of standardized appara-
tus and operational procedure is a disadvantage, but vari-
ations between TBP data reported by different laboratories
for the same sample are small because a close approach to
complete component separation is usually achieved. Mea-
surement of TBP data is more difficult than ASTM D 86 data
in terms of both time and cost. TBP and ASTM D 86 curves
for a kerosene sample are shown in Fig. 3.3 based on data
provided by Lenoir and Hipkin [ 1]. As shown in this figure
the IBP from TBP curve is less than the IBP from ASTM D
86 curve, while the FBP of TBP curve is higher than that of
ASTM curve. Therefore, the boiling range based on ASTM D
86 is less than the actual true boiling range. In TBP, the IBP
is the vapor temperature that is observed at the instant that
the first drop of condensate falls from the condenser.
3.1.1.3 Simulated Distillation by Gas Chromatography
Although ASTM D 86 test method is very simple and conve-
nient, it is not a consistent and reproducible method. For this
reason another method by gas chromatography (GC) is be-
ing recommended to present distillation data. A distillation
curve produced by GC is called a simulated distillation (SD)
and the method is described in ASTM D 2887 test method.
Simulated distillation method is simple, consistent, and
300 ...........................................................................................................................
. . . . . . . ASTM D86
250 TBP .-""
~'~ . . . . - " ~ 1 7 6 *" -'~
150
100
0 20 40 60 80 100
Vol% Vaporized
FIG. 3.3--ASTM D 86 and TBP curves for a
kerosene sample.
90 CHARACTERIZATION AND P R O P E R T I E S OF P E T R O L E U M FRACTIONS
250 ......................................................................................................................................................................
200 ....... ASTM D86
SD(ASTMD
150
100
50
__ i i
0 20 40
PercentVaporized
FIG. 3.4---Simulated and ASTM D 86 distilla-
tion curves for a petroleum fraction. (The per-
cent is in vol% for ASTM D 86 and is in wt% for
ASTM D 2887.)
reproducible and can represent the boiling range of a
petroleum mixture without any ambiguity. This method is ap-
plicable to petroleum fractions with a FBP of less than 538~
(1000~ and a boiling range of greater than 55~ (100~ and
having a vapor pressure sufficiently low to permit sampling at
ambient temperature. The ASTM D 2887 method is not appli-
cable to gasoline samples and the ASTM D 3710 test method
is recommended for such fractions. Distillation curves by SD
are presented in terms of boiling point versus wt% of mixture
vaporized because as described below in gas chromatogra-
phy composition is measured in terms of wt% or weight frac-
tion. Simulated distillation curves represent boiling points of
compounds in a petroleum mixture at atmospheric pressure;
however, as will be shown later SD curves are very close to ac-
tual boiling points shown by TBP curves. But these two types
of distillation data are not identical and conversion methods
should be used to convert SD to TBP curves. In comparison
with ASTM D 86, the IBP from a SD curve of a petroleum mix-
ture is less than IBP from ASTM D 86 curve, while the FBP
from SD curve is higher than the FBP from ASTM D 86 of the
same mixture (see Fig. 3.4). This is the same trend as that of
TBP curves in comparison with ASTM curves as was shown
in Fig. 3.3. A typical SD curve for a gas oil sample is shown
in Fig. 3.4. Note that in this figure the percent vaporized for
ASTM D 2887 (SD) is in wt% while for the ASTM D 86 curve
is in vol%.
The gas chromatography technique is a separation method
based on the volatility of the compounds in a mixture. The GC
is used for both generation of distillation curves as well as to
determine the composition of hydrocarbon gas or liquid mix-
tures, as will be discussed later in this chapter. For this reason
in this part we discuss the basic function of chromatography
techniques and elements of GC. In an analysis of a mixture by
a GC, the mixture is separated into its individual compounds
according to the relative attraction of the components for
a stationary and a mobile phase. Recent advances in chro-
matography make it possible to identify and separate com-
pounds with boiling points up to 750~ (1380~ A small fluid
sample (few microliters for liquid and 5 mL for gas samples)
is injected by a needle injector into a heated zone in which
the sample is vaporized and carried by a high-purity carrier
gas such as helium or nitrogen. The stationary phase is either
solid or liquid. A component that is more strongly attracted to
the mobile phase than to the stationary phase is swept along
with the mobile phase more rapidly than a component that is
~
more strongly attracted to the stationary phase. The mobile
phase can be a liquid phase as well; in this case the chro-
matography method is called liquid chromatography (LC).
The basic elements of a GC are a cylinder of carrier gas,
flow controller and pressure regulator, sample injector, col-
umn, detector, recorder, and thermostats for cylinder, col-
umn, and detector. The sample after injection enters a heated
oven where it enters the GC column (stationary phase). The
i i eluted components by the carrier gas called effluents enter
60 80 100 a detector where the concentration of each component may
be determined. The presence of a component in the carrier
gas leaving the column is continuously monitored through
an electric signal, which eventually is fed to a recorder for a
visual readout.
There are two types of columns, packed or capillary
columns, and two types of detectors, flame ionization detec-
tor or thermal conductivity detector. Packed columns have
inner diameters of 5-8 mm and length of 1-5 m. Column
and detector types depend on the nature of samples being
analyzed by the GC. The capillary columns are equivalent to
hundreds of theoretical equilibrium stages and can be used in
preference to packed columns. The inner diameter of capri-
lary columns is about 0.25-0.53 m m and their length is about
10-150 m. The stationary phase is coated on the inside wall
of columns. The flame ionization detector (FID) is highly sen-
sitive to all organic compounds (10 -12 g) but is not sensitive
to inorganic compounds and gases such as H20, CO2, N2,
and 02. The FID response is almost proportional to the mass
concentration of the ionized compound. Hydrogen of high
purity is used as the fuel for the FID. The thermal conduc-
tivity detector (TCD) is sensitive to almost all the compounds
but its sensitivity is less than that of FID. TCD is often used
for analysis of hydrocarbon gas mixtures containing nonhy-
drocarbon gases. The retention time is the amount of time re-
quired for a given component spent inside the column from
its entrance until its emergence from the column in the efflu-
ent. Each component has a certain retention time depending
on the structure of compound, type of column and station-
ary phase, flow rate of mobile phase, length, and tempera-
ture of column. More volatile compounds with lower boil-
ing points have lower retention times. Detector response is
measured in millivolts by electric devices. The written record
obtained from a chromatographic analysis is called a chro-
matograph. Usually the time is the abscissa (x axis) and mV
is the ordinate (y axis). A typical chromatograph obtained
to analyze a naphtha sample from a Kuwaiti crude is shown
in Fig. 3.5. Each peak corresponds to a specific compound.
Qualitative analysis with GC is done by comparing retention
times of sample components with retention times of reference
compounds (standard sample) developed under identical ex-
perimental conditions. With proper flow rate and tempera-
ture, the retention time can he reproduced within I%. Ev-
ely component has only one retention time; however, compo-
nents having the same boiling point or volatility but different
molecular structure cannot be identified through GC analysis.
In Fig. 3.5, compounds with higher retention time (x coordi-
nate) have higher boiling points and the actual boiling point
or the compound can be determined by comparing the peak
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S 91

q~

J i

+
2~Q
i
i
J

!
+ti
r'o z~ ~ ~- ................../~ ~ ,45

FIG. 3.5~A typical chromatograph for a Kuwaiti naphtha sample.

with the s i m i l a r p e a k of a k n o w n c o m p o u n d w i t h a k n o w n cent is the s a m e as c o m p o s i t i o n in wt% with boiling p o i n t s of

boiling point. In the quantitative analysis of a mixture, it can
be s h o w n t h a t the a r e a u n d e r a p a r t i c u l a r c o m p o n e n t p e a k (as
s h o w n in Fig. 3.5) is directly p r o p o r t i o n a l to the total a m o u n t
(mass) of the c o m p o n e n t r e a c h i n g the detector. The a m o u n t
r e a c h i n g the d e t e c t o r is also p r o p o r t i o n a l to the concentra-
tion (weight p e r c e n t or weight fraction) of the c o m p o n e n t
in the s a m p l e injected. The p r o p o r t i o n a l i t y c o n s t a n t is deter-
m i n e d with the aid of s t a n d a r d s c o n t a i n i n g a k n o w n a m o u n t
of the s a m p l e c o m p o n e n t . M o d e m GCs are e q u i p p e d w i t h a
c o m p u t e r t h a t directly m e a s u r e s the a r e a s u n d e r each p e a k
a n d c o m p o s i t i o n can be directly d e t e r m i n e d f r o m the com-
p u t e r printout. A p r i n t o u t for the c h r o m a t o g r a p h of Fig. 3.5
is shown in Table 3.1 for the n a p h t h a sample. The a r e a per-
c o r r e s p o n d i n g c o m p o n e n t s . In analysis of s a m p l e s b y a GC,
the c o m p o s i t i o n i s always d e t e r m i n e d in wt% a n d n o t in vol%
o r fraction. F o r this r e a s o n the o u t p u t of a GC analysis for a
s i m u l a t e d distillation is a curve of t e m p e r a t u r e (boiling point)
versus wt% of m i x t u r e vaporized, as can be seen in Fig. 3.4.
F u r t h e r i n f o r m a t i o n for use of GC for s i m u l a t e d distillation
up to 750~ is p r o v i d e d b y Curvers a n d van d e n Engel [3].
A typical GC for m e a s u r e m e n t of boiling p o i n t of p e t r o l e u m
p r o d u c t s is s h o w n in Fig. 3.6.
3.1.1.4 E q u i l i b r i u m Flash Vaporization
E q u i l i b r i u m flash v a p o r i z a t i o n (EFV) is the least i m p o r t a n t
type of distillation curve a n d is very difficult to m e a s u r e . It is

TABLE 3.1--Calculation of composition of a naphtha sample with GC chromatograph shown in Fig. 3.4.
No. Name Time, min Area Area % Tb,~C
1 n-Hexane 11.16 1442160 3.92 68.7
2 Benzene 14.64 675785 1.84 80.1
3 Cyclohexane 15.55 3827351 10.40 80.7
4 n-Heptane 18.90 5936159 16.14 98.4
5 2,2,3-Trimethyl-pentane 20.38 8160051 22.18 109.8
6 Toluene 27.53 8955969 24.34 110.6
7 Ethylbenzene 42.79 1678625 4.56 136.2
8 p-Xylene 45.02 4714426 12.82 138.4
9 o-Xylene 49.21 1397253 3.8 144.4
Total 36787780 100
92 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

FIG. 3.6--A GC for measurement of boiling point of products (courtesy of KISR).

p r e s e n t e d in t e r m s of the t e m p e r a t u r e versus vol% vaporized. 3.1.1.5 Distillation at Reduced Pressures

It involves a series of e x p e r i m e n t s at c o n s t a n t a t m o s p h e r i c
p r e s s u r e w i t h total v a p o r in e q u i l i b r i u m w i t h the unvapor-
ized liquid. In fact to d e t e r m i n e each p o i n t on the EFV curve
one e x p e r i m e n t is required. To have a full shape of an EFV
curve at least five t e m p e r a t u r e s at 10, 30, 50, 70, a n d 90 vol%
v a p o r i z e d are required. EFV distillation curves are useful in
the design a n d o p e r a t i o n of o v e r h e a d p a r t i a l c o n d e n s e r s a n d
b o t t o m reboilers since the EFV t e m p e r a t u r e s r e p r e s e n t actual
e q u i l i b r i u m t e m p e r a t u r e s . In c o n t r a s t w i t h TBR the EFV ini-
tial t e m p e r a t u r e of a mixture is greater t h a n the IBP of ASTM
D 86 curve, while the FBP f r o m a EFV curve is lower t h a n the
FBP from the TBP curve for the s a m e mixture. EFV curves at
p r e s s u r e s above a t m o s p h e r i c u p to p r e s s u r e s of 15 b a r m a y
also b e useful for design a n d o p e r a t i o n of v a p o r i z i n g o r con-
densing vessels u n d e r pressure.
A t m o s p h e r i c distillation curves p r e s e n t boiling p o i n t s of
p r o d u c t s from a n a t m o s p h e r i c distillation column. F o r p r o d -
ucts such as heavy gas oils t h a t c o n t a i n heavy c o m p o u n d s a n d
m a y u n d e r g o a cracking p r o c e s s d u r i n g v a p o r i z a t i o n at a t m o -
spheric pressure, distillation d a t a are m e a s u r e d at r e d u c e d
p r e s s u r e s from I to 760 m m Hg. The e x p e r i m e n t a l p r o c e d u r e
is d e s c r i b e d in ASTM D 1160 test m e t h o d (see Fig. 3.7). Distil-
lation of heavy p e t r o l e u m fractions is n o r m a l l y p r e s e n t e d at
1, 2, 10, or 50 m m H g . Both a m a n u a l a n d a n a u t o m a t i c
m e t h o d are specified. The t e m p e r a t u r e of the v a p o r s h o u l d
not exceed 400~ (750~ ASTM D 1160 distillation d a t a are
m e a s u r e d m o r e accurately t h a n ASTM D 86 since it is con-
d u c t e d at low pressure. F o r this r e a s o n ASTM D 1 I60 curves
a r e closer to TBP curves at the s a m e p r e s s u r e base. Con-
version of distillation d a t a f r o m low p r e s s u r e to equivalent

FIG. 3.7--An apparatus for experimental measurement of boiling point

at reduced pressures by ASTM D 1160 test method (courtesy of KISR).
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S
atmospheric boiling points are given in the ASTM Manual [4]
and will be discussed later in this chapter.
3.1.2 Density, Specific Gravity, and API Gravity
Specific gravity (SG) or relative density and the API gravity are
defined in Section 2.1.3 and for pure hydrocarbons are given
in Tables 2.1 and 2.2. Aromatic oils are denser than paraffinic
oils. Once specific gravity is known, the API gravity can be de-
termined by Eq. (2.4), which corresponds to the ASTM D 287
method. The standard temperature to measure the specific
gravity is 15.56~ (60~ however, absolute density is usually
reported at 20~ Specific gravity or density for a petroleum
mixture is a bulk property that can be directly measured for
the mixture. Specific gravity is a property that indicates the
quality of a petroleum product and, as was shown in Chapter
2, is a useful property to estimate various physical proper-
ties of petroleum fluids. A standard test method for density
and specific gravity of liquid petroleum products and distil-
lates in the range of 15-35~ through use of a digital density
meter is described in ASTM D 4052 method [4]. The appa-
ratus must be calibrated at each temperature and this test
method is equivalent to ISO 12185 and IP 365 methods. An-
other method using a hydrometer is described under ASTM D
1298 test method. Hydrometer is a glass float with lead ballast
that is floated in the liquid. The level at which hydrometer is
floating in the liquid is proportional to the specific gravity
of the liquid. Through graduation of the hydrometer the spe-
cific gravity can be read directly from the stalk of hydrometer.
This method is simpler than the ASTM 4052 method but is
less accurate. The French standard procedure for measuring
density by hydrometer is described under NFT 60-101 test
method. With some hydrometers densities with accuracy of
0.0005 g/mL can be measured. A digital density meter model
DMA 48 from PARA (Austria) is shown in Fig. 3.8.
FIG. 3.8--PARA model DMA 48 digital density meter (courtesy of Chemical Engineering
Department at Kuwait University).
93
3.1.3 Molecular Weight
Molecular weight is another bulk property that is indicative of
molecular size and structure. This is an important property
that usually laboratories do not measure and fail to report
when reporting various properties of petroleum fractions.
This is perhaps due to the low accuracy in the measurement
of the molecular weight of petroleum fractions, especially for
heavy fractions. However, it should be realized that experi-
mental uncertainty in reported values of molecular weight
is less than the errors associated with predictive methods for
this very useful parameter. Since petroleum fractions are mix-
tures of hydrocarbon compounds, mixture molecular weight
is defined as an average value called number average molec-
ular weight or simply molecular weight of the mixture and it
is calculated as follows:
(3.1) M = ~x4Mi
i
where xi and M/are the mole fraction and molecular weight of
component i, respectively. Molecular weight of the mixture,
M, represents the ratio of total mass of the mixture to the total
moles in the mixture. Exact knowledge of molecular weight
of a mixture requires exact composition of all compounds in
the mixture. For petroleum fractions such exact knowledge is
not available due to the large number of components present
in the mixture. Therefore, experimental measurement of mix-
ture molecular weight is needed in lieu of exact composition
of all compounds in the mixture.
There are three methods that are widely used to measure
the molecular weight of various petroleum fractions. These
are cryoscopy, the vapor pressure method, and the size ex-
clusion chromatography (SEe) method. For heavy petroleum
fractions and asphaltenic compounds the SEC method is
commonly used to measure distribution of molecular weight
94 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
FIG. 3.9--A GPC from waters model 150-C plus (courtesy of Chemical Engineering
Department at Kuwait University).
in the fraction. The SEC method is mainly used to determine
molecular weights of polymers in the range of 2000 to 2 • 106.
This method is also called gel permeation chromatography
(GPC) and is described in the ASTM D 5296 test method. In
the GPC method, by comparing the elution time of a sample
with that of a reference solution the molecular weight of the
sample can be determined. A GPC instrument is shown in
Fig. 3.9. The SEC experiment is usually performed for heavy
residues and asphaltenes in crude oils and gives the wt% of
various constituents versus molecular weight as will be dis-
cussed in Chapter 4.
The vapor pressure method is based on the measurement of
the difference between vapor pressure of sample and that of
a known reference solvent with a vapor pressure greater than
that of the sample. A solution of about 1 g of sample in 25 mL
of the reference solvent is prepared. This solution, which has
vapor pressure less than that of the solvent, tends to condense
the vapors of solvent on the sample thus creating a temper-
ature difference which is measured by two thermistors. The
molarity of the solution is calculated using calibrated curves.
This method is described by the ASTM D 2503 test method
and is applicable to oils with an initial boiling point greater
than 220~ [5]. A typical experimental error and uncertainty
in measuring the molecular weight is about 5%.
The third and most widely used method of determining
the molecular weight of an unknown petroleum mixture is
by the cryoscopy method, which is based on freezing point
depression. The freezing point of a solution is a measure of
the solution's concentration. As the concentration of the so-
lute increases, the freezing point of the solution will be lower.
The relation between freezing point depression and concen-
tration is linear. For organic hydrocarbons, benzene is usually
used as the solvent. Special care should be taken when work-
ing with benzene [6]. Calibration curves can be prepared by
measuring the freezing points of different solute concentra-
tions with a known solute and a known solvent. A cryoscope
can measure the freezing point depression with an accuracy
of about 0.001 ~ The relation to obtain molecular weight of
a sample is [6]
1000 x Kf x ml
(3.2) M=
AT x m2
where Kf is molal freezing point depression constant of the
solvent and is about 5.12~ AT is the freezing point de-
pression and the reading from the cryoscope, rnl is the mass
of solute and m2 is the mass of solvent both in grams. It gener-
ally consists of refrigerator, thermometer and the apparatus
to hold the sample. A cryoscope is shown in Fig. 3.10.
3.1.4 Refractive Index
Refractive index or refractivity is defined in Section 2.1.4 and
its values at 20~ for pure hydrocarbons are given in Table
2.1. Refractive indexes of hydrocarbons vary from 1.35 to 1.6;
however, aromatics have refractive index values greater than
naphthenes, which in turn have refractive indexes greater
than paraffins. Paraffinic oils have lower refractive index val-
ues. It was shown in Chapter 2 that refractive index is a useful
parameter to characterize hydrocarbon systems and, as will
be seen later in this chapter, it is needed to estimate the com-
position of undefined petroleum fractions. Refractive index is
the ratio of the speed of light in a vacuum to that of a medium.
In a medium, the speed of light depends on the wavelength
and temperature. For this reason refractive index is usually
measured and reported at 20~ with the D line sodium light.
For mixtures, refractive index is a bulk property that can
be easily and accurately measured by an instrument called
a refractometer Refractive index can be measured by digital
refractometers with a precision of 4-0.0001 and temperature
precision of 4-0. I~ The amount of sample required to mea-
sure refractive index is very small and ASTM D 1218 provides
a test method for clear hydrocarbons with values of refractive
indexes in the range of 1.33-1.5 and the temperature range
of 20-30~ In the ASTM D 1218 test method the Bausch and
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS 95

FIG. 3.10~Model 5009 wide range cryoscope to measure molecular weight (courtesy of
Chemical Engineering Department at Kuwait University).

Lomb refractometer is used. Refractive index of viscous oils

with values up to 1.6 can be measured by the ASTM D 1747
test method. Samples must have clear color to measure their
refractive index; however, for darker and more viscous sam-
ples in which the actual refractive index value is outside the
range of application of refractometer, samples can be diluted
by a light solvent and refractive index of the solution should be
measured. From the composition of the solution and refrac-
tive index of pure solvent and that of the solution, refractive
index of viscous samples can be determined. A Model Abbe re-
fractometer (Leica) is shown in Fig. 3.11. This refractometer
measures refractive index of liquids within the temperature
range of - 2 0 to 100~ with temperature accuracy of •176
Because of simplicity and importance of refractive index it
would be extremely useful if laboratories measure and report
its value at 20~ for a petroleum product, especially if the
composition of the mixture is not reported.

3.1.5 Compositional Analysis

Petroleum fractions are mixtures of many different types of
hydrocarbon compounds. A petroleum mixture is well defined
if the composition and structure of all compounds present in
the mixture are known. Because of the diversity and number
of constituents of a petroleum mixture, the determination of
such exact composition is nearly impossible. Generally, hy-
drocarbons can be identified by their carbon number or by
their molecular type. Carbon number distribution may be
determined from fractionation by distillation or by molec-
ular weight distribution as discussed earlier in this section.
However, for narrow boiling range petroleum products and
petroleum cuts in which the carbon number range is quite
limited, knowledge of molecular type of compounds is very
important. As will be seen later, properties of petroleum frac-
tions with detailed compositional analysis can be estimated
with a higher degree of accuracy than for undefined fractions. FIG. 3.11--Leica made Abbe refractometer (courtesy of
After distillation data, molecular type composition is the Chemical Engineering Department at Kuwait University).
most important characteristic of petroleum fractions. In this
96 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

section various types of composition of petroleum fractions Feedstock

and different methods of their measurement are presented. n-Heptane (or n-Pentane)
3.1.5.1 Types of Composition 1
Based on the nature of petroleum mixture, there are sev-
eral ways to express the composition of a petroleum mixture.
Some of the most important types of composition are given
below:
1
Asphaltenes
t
DeasphaltenedOil

9 PONA (paraffins, olefins, naphthenes, and aromatics) Acetone

9 PNA (paraffins, naphthenes, and aromatics) /

9 PIONA (paraffins, isoparaffins, olefins, naphthenes, and aro- ]

matics)
9 SARA (saturates, aromatics, resins, and asphalthenes)
9 Elemental analysis (C, H, S, N, O)
1
Resins
1
Oils
Since most petroleum fractions are free of olefins, the hydro-
Dimethylforrnamide
carbon types can be expressed in terms of only PINA and if
paraffins and isoparaffins are combined a fraction is simply
expressed in terms of PNA composition. This type of anal-
ysis is useful for light and narrow boiling range petroleum
products such as distillates from atmospheric crude dis-
tillation units. But the SARA analysis is useful for heavy
1
Aromatics
1
Saturates

petroleum fractions, residues, and fossil fuels (i.e., coal liq-
uids), which have high contents of aromatics, resins, and as-
phaltenes. The elemental analysis gives information on hy-
drogen and sulfur contents as well as C/H ratio, which are
indicative of the quality of petroleum products.
3.1.5.2 Analytical Instruments
Generally three methods may be used to analyze petroleum
fractions. These are
9 separation by solvents
9 chromatography methods
9 spectroscopic methods
The method of separation by solvents is based on solubil-
ity of some compounds in a mixture in a particular solvent.
The remaining insoluble compounds may be in a solid or an-
other immiscible liquid phase. This method is particularly
useful for heavy petroleum fractions and residues contain-
ing asphahenes, resins, and saturate hydrocarbons. The de-
gree of solubility of a compound in a solvent depends on the
chemical structure of both the solute and the solvent. If the
two structures are similar there is a greater degree of solubil-
ity. For example, high-molecular-weight asphaltenes are not
soluble in a low-molecular-weight paraffinic solvent such as
n-heptane. Therefore, if n-heptane is added to a heavy oil, as-
phaltenes precipitate while the other constituents form a solu-
FIG. 3.12--An all-solvent fractionation procedure. Reprinted
from Ref. [7], p. 267, by courtesy of Marcel Dekker, Inc.
removed from the product [7]. ASTM [4] provides several
methods based on solvent separation to determine amounts
of asphaltenes. In ASTM D 2007 test method n-pentane is used
as the solvent, while in ASTM D 4124 asphaltene is separated
by n-heptane. Schematics of these test methods are shown in
Figs. 3.13 and 3.14, respectively, as given by Speight [7]. As-
phaltenes are soluble in liquids with a surface tension above
25 dyne/cm such as pyridine, carbon disulfide, carbon tetra-
chloride, and benzene [7].
The principle of separation by chromatography technique
was described in Section 3.1.1.3. If the mobile phase is gas
the instrument is called a gas chromatograph (GC), while for
Feedstock/
[ n-Pentane
/
|
Asphaltenes Deasphaltened0il
Clay
/

ble solution with the solvent. If solvent is changed to propane, |

because of the greater difference between the structure of
the solvent and the high-molecular-weight asphaltenes, more
asphaltenic compounds precipitate. Similarly if acetone is Resins Oils
added to a deasphalted oil (DAO), resins precipitate while
low-molecular-weight hydrocarbons remain soluble in ace- ~l SilicaGel
tone. In Fig. 3.12 an all-solvent fractions procedure is shown
for SARA analysis [7].
One of the disadvantages of the all-solvent separation tech-
nique is that in some instances a very low temperature (0 to
- 10 ~C) is required, which causes inconvenience in laboratory
operation. Another difficulty is that in m a n y cases large vol-
l
Aromatics
1
Saturates
umes of solvent may be required and solvents must have suffi- FIG. 3.13---The ASTM D 2007 procedure. Reprinted from
ciently low boiling point so that the solvent can be completely Ref. [7], p. 280, by courtesy of Marcel Dekker, inc.
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS 97

Feedstock stream may be analyzed every 20 min and automatic adjust-

ment can be made to the refinery unit. In crude assay anal-
n-Heptane
ysis distillation is being replaced by chromatography tech-
niques. The LC method is used for less volatile mixtures such

l
Asphaltenes
1
DeasphaltenedOil
as heavy petroleum fractions and residues. Use of LC for sepa-
ration of saturated and aromatic hydrocarbons is described in
ASTM D 2549 test method. Various forms of chromatography
techniques have been applied to a wide range of petroleum
~ Alumina products for analysis, such as PONA, PIONA, PNA, and SARA.
One of the most useful types of liquid chromatography is
high performance liquid chromatography (HPLC), which can
be used to identify different types of hydrocarbon groups.
Polar Aromatics Oils One particular application of HPLC is to identify asphaltene
(Resins) ~ Alumina and resin type constituents in nonvolatile feedstocks such as
residua. Total time required to analyze a sample by HPLC is
just a few minutes. One of the main advantages of HPLC is

l
NaphtheneAromatics Saturates
(Aromatics)
FIG. 3.14.--Separation of asphaltenes and resins from
petroleum fractions. Reprinted from Ref. [7], p. 281, by cour-
tesy of Marcel Dekker, Inc.
liquid mobile phase it is called a liquid chromatograph (LC).
As discussed earlier, components can be separated by their
boiling points through GC analysis. In advanced petroleum
refineries automatic online GCs are used for continuous anal-
ysis of various streams to control the quality of products. A
that the boiling range of sample is immaterial. A HPLC ana-
lyzer is shown in Fig. 3.15.
The accuracy of chromatography techniques mainly
depends on the type of detector used [7]. In Section 3.1.1.3,
flame ionization (FID) and thermal conductivity (TCD) detec-
tors are described, which are widely used in GC. For LC the
most common detectors are refractive index detector (RID)
and wavelength UV (ultraviolet) detector. UV spectroscopy is
particularly useful to identify the types of aromatics in asphal-
tenic fractions. Another spectroscopy method is conventional
infrared (IR) spectroscopy, which yields information about
the functional features of various petroleum constituents.
For example, IR spectroscopy will aid in the identification of
N--H and O--H functions and the nature of polymethylene

FIG. 3.15--A HPLC instrument (courtesy of Chemical Engineering Department at Kuwait

University).
98 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
FIG. 3,16--HP made GC-MS model 5890 Series II. (courtesy of Chemical Engineering
Department at Kuwait University),
chains (C--H) and the nature of any polynuclear aromatic
systems [7].
Another type of analysis of petroleum fractions to iden-
tify molecular groups is by spectrometric methods such as
mass spectrometry (MS). In general, there is a difference be-
tween spectroscopy and spectrometry methods although in
some references this difference is not acknowledged. Spec-
troscopy refers to the techniques where the molecules are ex-
cited by various sources, such as UV and IR, to return to their
normal state. Spectrometry refers to the techniques where
the molecules are actually ionized and fragmented. Evolution
of spectroscopic methods comes after chromatography tech-
niques, nonetheless, and in recent decades they have received
considerable attention. While volatile and light petroleum
products can be analyzed by gas chromatography, heavier
and nonvolatile compounds can be analyzed and identified by
spectrometric methods. One of the most important types of
spectrometry techniques in analysis of petroleum fractions is
mass spectrometry (MS). In this method, masses of molecular
and atomic components that are ionized in the gaseous state
by collision with electrons are measured. The advantage of
MS over other spectrometric methods is its high reproducibil-
ity of quantitative analysis and information on molecular type
in complex mixtures. Mass spectrometry can provide the most
detailed quantitative and qualitative information about the
atomic and molecular composition of organic and inorganic
compounds. However, use of MS is limited to organic com-
pounds that are stable up to 300~ (570~ At higher tem-
peratures thermal decomposition may occur and the anal-
ysis will be biased [7]. Through MS analysis, hydrocarbons
of similar boiling points can be identified. In the MS analy-
sis, molecular weight, chemical formula of hydrocarbons, and
their amounts can be determined. The most powerful instru-
ment to analyze petroleum distillates is the combination of
a GC and an MS called GC-MS instrument, which separates
compounds both through boiling point and molecular weight.
For heavy petroleum fractions containing high-boiling-point
compounds an integrated LC-MS unit may be suitable for
analysis of mixtures; however, use of LC-MS is more difficult
than GC-MS because in LC-MS solvent must be removed
from the elute before it can be analyzed by MS. A GC-MS
instrument from Hewlett Packard (HP) is shown in Fig. 3.16
Another type of separation is by SEC or GPC, which can
be used to determine molecular weight distribution of heavy
petroleum fractions as discussed in Section 3.1.3. Fractions
are separated according to their size and molecular weight
and the method is particularly useful to determine the amount
of asphaltenes. Asphaltenes are polar multiring aromatic
compounds with molecular weight above 1000 (see Fig. 1.2).
It is assumed that in this molecular weight range only aro-
matics are present in a petroleum fraction [8].
3.1.5.3 PNA Analysis
As determination of the exact composition of a petroleum
fraction is nearly an impossible task, for narrow boiling range
petroleum fractions and products a useful type of compo-
sitional analysis is to determine the amounts of paraffins
(P), naphthenes (N), and aromatics (A). As mentioned be-
fore, most petroleum products are olefin free and PNA anal-
ysis provides a good knowledge of molecular type of mixture
constituents. However, some analyzers give the amount of
isoparaffins and olefins as well. These analyzers are called
PIONA analyzer, and a Chrompack Model 940 PIONA ana-
lyzer is shown in Fig. 3.17. An output of this type of analyzer
is similar to the GC output; however, it directly gives wt% of
n-paraffins, isoparaffins, olefins, naphthenes, and aromatics.
The composition is expressed in wt%, which can be converted
to mole, weight, and volume fractions as will be shown later
in this chapter.
3.1.5.4 Elemental Analysis
The main elements present in a petroleum fraction are car-
bon (C), hydrogen (H), nitrogen (N), oxygen (O), and sulfur
(S). The most valuable information from elemental analysis
 3. CHARACTERIZATION OF P E T R O L E U M FRACTIONS
FIG. 3,17mA chrompack model 940 PIONA analyzer (courtesy of Chemical Engineering
Department at Kuwait University),
that can be obtained is on the C/H ratio and sulfur content
of a petroleum mixture from which one can determine the
quality of oil. As boiling points of fractions increase or their
API gravity decrease the amount of C/H ratio, sulfur content,
nitrogen content, and the metallic constituents increase, sig-
nifying a reduction in the quality of an oil. Sulfur content of
very heavy fractions can reach 6-8% and the nitrogen content
can reach 2.0-2.5 wt%. There are specific methods to measure
these elements individually. However, instruments do exist
that measure these elements all together; these are called ele-
mental analyzers. One of these apparatuses is CHN analyzers
in which there is a simultaneous combustion in pure oxygen
at 1000~ Carbon is reduced to CO2, H is reduced to H20,
and N is converted to nitrogen oxides. Nitrogen oxides are
then reduced over copper at 650~ to nitrogen by eliminating
oxygen. A mixture of CO2, H/O, and N2 is separated by gas
chromatography with TCD. In a similar fashion, sulfur is ox-
idized to SO2 and is detected by TCD after detection of CO2,
Na, and H20. Oxygen is determined by passing it over carbon
at high temperature and converted to CO, which is measured
by a GC [5]. ASTM test methods for elemental analysis of
petroleum products and crude oils include hydrogen content
(ASTM D 1018, D 3178, D 3343), nitrogen content (ASTM
D 3179, D 3228, D 3431), and sulfur content (ASTM D 129/IP
61, D 1266/IP 107, D 1552, D 4045). An elemental analyzer
Model CHNS-932 (Leco Corp., St. Joseph, MI, USA) is shown
in Fig. 3.18. In this analyzer, the CO2, H20, and SO2 are
detected by infrared detector (IRD) while N2 is determined
by the TCD method.
Another group of heteroatoms found in petroleum mix-
tures are metallic constituents. The amount of these met-
als are in the range of few hundreds to thousand ppm and
their amounts increase with increase in boiling points or de-
crease in the API gravity of oil. Even a small amount of these
metals, particularly nickel, vanadium, iron, and copper, in
the feedstocks for catalytic cracking have negative effects on
the activity of catalysts and result in increased coke forma-
99
tion. Metallic constituents are associated with heavy com-
pounds and mainly appear in residues. There is no general
method to determine the composition of all metals at once
but ASTM [4] provides test methods for determination of
various metallic constituents (i.e., ASTM D 1026, D 1262,
D 1318, D 1368, D 1548). Another method is to burn the oil
sample in which metallic compounds appear in inorganic
ashes. The ash should be digested by an acid and the so-
lution is examined for metal species by atomic absorption
spectroscopy [7].
3.1.6 Viscosity
Absolute and kinematic viscosities are defined in Section
2.1.8 and experimental data for the kinematic viscosity of
some pure hydrocarbons are given in Table 2.2. Viscosity of
petroleum fractions increase with a decrease in the API grav-
ity and for residues and heavy oils with the API gravity of
less than 10 (specific gravity of above 1), viscosity varies from
several thousands to several million poises. Viscosity is a bulk
property that can be measured for all types of petroleum frac-
tions in liquid form. Kinematic viscosity is a useful character-
ization parameter for heavy fractions in which boiling point
data are not available due to thermal decomposition during
distillation. Not only is viscosity an important physical prop-
erty, but it is a parameter that can be used to estimate other
physical properties as well as the composition and quality of
undefined petroleum fractions as shown later in this chap-
ter. Since viscosity varies with temperature, values of viscos-
ity must be reported with specified temperature. Generally,
kinematic viscosity of petroleum fractions are measured at
standard temperatures of 37.8~ (100~ and 98.9~ (210~
However, for very heavy fractions viscosity is reported at tem-
peratures above 38~ i.e., 50~ (122~ or 60~ (140~ When
viscosity at two temperatures are reported from the method
of Section 2.7 one can obtain the viscosity at other tempera-
tures. Measurement of viscosity is easy but the method and
100 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
FIG. 3.18--Leco made CHNS-932 model elemental analyzer (courtesy of Chemical Engi-
neering Department at Kuwait University).
the instrument depend on the type of sample. For Newto-
nian and high-shear fluids such as engine oils, viscosity can
be measured by a capillary U-type viscometer. An example of
such viscometers is the Cannon-Fenske viscometer. The test
method is described in ASTM D 445, which is equivalent to
ISO 3104 method, and kinematic viscosity is measured at tem-
peratures from 15 to 100~ (~60 to 210~ In this method,
repeatability and reproducibility are 0.35 and 0.7%, respec-
tively [5]. Another type of viscometer is a rotary viscometer,
which is used for a wide range of shear rates, especially for
low shear rate and viscous fluids such as lubricants and heavy
petroleum fractions. In these viscometers, fluid is placed be-
tween two surfaces, one is fixed and the other one is rotating.
In these viscometers absolute viscosity can be measured and
an example of such viscometers is the Brookfield viscome-
ter. Details of measurement and prediction of viscosity of
petroleum fractions are given in Chapter 8. As the viscosity
of petroleum fractions, especially the heavy oils, is one of the
most difficult properties to estimate, its experimental value is
highly useful and desirable.
3.2 PREDICTION AND CONVERSION
OF DISTILLATION DATA
Various distillation curves are introduced in Section 3.1.1.
For simplicity ASTM is used to refer to ASTM D 86 distilla-
tion curve, similarly TBP, SD, and EFV refer to true boiling
point, simulated distillation (ASTM D 2887), and equilibrium
flash vaporization, respectively. Petroleum fractions have a
range of boiling points. To use the correlations introduced
in Chapter 2, a single value for boiling point is required. For
this reason there is a need for the definition of an average
boiling point or a characteristic boiling point based on a dis-
tillation curve. Availability of one type of distillation curve
for simplicity in experimental measurement and the need for
another type for its application requires conversion methods
between various distillation curves. The tedious procedures
necessary to obtain experimental EFV data have given im-
petus to the development of correlations for predicting EFV
data from the analytical ASTM and TBP distillations. Sim-
ulated distillation by gas chromatography appears to be the
most simple, reproducible, and consistent method to describe
the boiling range of a hydrocarbon fraction unambiguously.
TBP is the most useful distillation curve, while available data
might be ASTM D 86, ASTM D 2887, or ASTM D 1160 distilla-
tion curves. ASTM [4] has accepted this technique as a tenta-
tive method for the "Determination of Boiling Range Distribu-
tion of Petroleum Fractions by Gas Chromatography" (ASTM
D 2887). In most cases distillation data are reported in terms
of ASTM D 86 or SD. In this section methods of calculation of
average boiling points, interconversion of various distillation
curves, and prediction of complete distillation curves from a
limited data are presented.
3.2.1 Average Boiling Points
Boiling points of petroleum fractions are presented by distil-
lation curves such as ASTM or TBR However, in prediction of
physical properties and characterization of hydrocarbon mix-
tures a single characteristic boiling point is required. Gener-
ally an average boiling point for a fraction is defined to de-
termine the single characterizing boiling point. There are five
average boiling points defined by the following equations [9].
Three of these average boiling points are VABP (volume aver-
age boiling point), MABP (molal average boiling point) and
WABP (weight average boiling point), defined for a mixture
of n components as
tt
(3.3) ABP = ~-~ x4Tb~
i=1

w h e r e ABP is the VABP, MABP, o r WABP a n d xi is the corre-
s p o n d i n g volume, mole, o r weight fraction of c o m p o n e n t i.
Tbi is the n o r m a l boiling p o i n t of c o m p o n e n t i in kelvin. Two
o t h e r average boiling points are CABP (cubic average boiling
point) a n d MeABP ( m e a n average boiling point) defined as
(3.4) CABP = x~ (1.8Tbi -- 459.67) 1/3
MABP + CABP
(3.5) MeABP =
2 CABP, a n d MeABP for this fraction.
w h e r e T~ in Eq. (3.4) is in kelvin. The conversion factors in
Eq. (3.4) c o m e from the fact t h a t the original definition of
CABP is in degrees Fahrenheit. F o r p e t r o l e u m fractions in
w h i c h volume, weight, o r m o l e fractions of c o m p o n e n t s are
n o t known, the average boiling p o i n t s are calculated t h r o u g h
ASTM D 86 distillation curve as
T10 + T30 + T50 + T70 + T90
(3.6) VABP =
5
w h e r e 1"10, T30, Ts0, I"70, a n d T90 are ASTM t e m p e r a t u r e s at
10, 30, 50, 70, a n d 90 vol% distilled. ASTM distillation curves
c a n be c h a r a c t e r i z e d by the m a g n i t u d e s of t e m p e r a t u r e s a n d
overall slope of the curve. A p a r a m e t e r t h a t a p p r o x i m a t e l y
c h a r a c t e r i z e s slope o f a distillation curve is the slope o f a
l i n e a r line between 10 and 90% points. This slope shown by
SL is defined as
T90 - Tl0
(3.7) SL -
80
w h e r e T10 a n d T90 are the ASTM D 86 t e m p e r a t u r e s at 10 a n d
90% of v o l u m e vaporized. The 10-90 slope, SL, in s o m e refer-
ences is referred to as the Engler slope a n d is indicative of a
variety of c o m p o u n d s in a p e t r o l e u m fraction. W h e n the boil-
ing points of c o m p o u n d s are n e a r each o t h e r the value of SL
a n d the boiling range of the fraction are low. F o r p e t r o l e u m
fractions, WABP, MABP, CABP, a n d MeABP are c o r r e l a t e d
t h r o u g h an e m p i r i c a l plot to VAPB a n d SL in C h a p t e r 2 of the
API-TDB [2]. Analytical correlations b a s e d on the API plot
were developed b y Zhou [10] for use in a digital computer.
F o r heavy fractions a n d v a c u u m distillates in w h i c h distilla-
tion d a t a by ASTM D i 160 are available, they should first be
converted to ASTM D 86 a n d then average boiling p o i n t s are
calculated. Analytical correlations for e s t i m a t i o n of average
boiling p o i n t s are given by the following equations in t e r m s
of VABP a n d SL [2, 10].
(3.8) ABP = VABP - AT
l n ( - A T w ) = -3.64991 - 0.02706(VABP - 273.15) 0.6667
(3.9) + 5.163875SL ~
ln(ATM) = - 1.15158 - 0.01181 (VABP - 273.15) 0"6667
(3.10) + 3.70612SL 0'333
ln(ATc) = - 0 . 8 2 3 6 8 - 0.08997(VABP - 273.15) 0.45
(3.1 I) + 2.456791SL ~
ln(ATue) = - 1.53181 - 0.0128(VABP - 273.15) 0.6667
(3.12) + 3.646064SL ~
w h e r e ABP is an average boiling p o i n t such as WABP, MABP,
CABP, o r MeABP a n d AT is the c o r r e s p o n d i n g c o r r e c t i o n
t e m p e r a t u r e for each ABP. All t e m p e r a t u r e s are in kelvin.
3. C H A R A C T E R I Z A T I O N OF PETROLEUM FRACTIONS 101
VABP a n d SL are defined in Eqs. (3.6) a n d (3.7). Once AT
is calculated for each case, c o r r e s p o n d i n g ABP is calculated
from Eq. (3.8). E q u a t i o n s (3.8)-(3.12) calculate values of var-
ious ABP very close to those o b t a i n e d from e m p i r i c a l plot in
the API-TDB [2]. The following example shows a p p l i c a t i o n of
these equations in calculation of various ABE
+ 255.37
E x a m p l e 3 . 1 - - A low boiling n a p h t h a has the ASTM D 86 tem-
p e r a t u r e s of 77.8, 107.8, 126.7, 155, a n d 184.4~ at 10, 30, 50,
70, a n d 90 vol% distilled [11]. Calculate VABE WABE MABP,
S o l u t i o n - - F r o m Eqs. (3.6) a n d (3.7) VABP a n d SL are cal-
culated as follows: VABP = (77.8 + 107.8 + 126.7 + 155 +
184.4)/5 = 130.3~ = 403.5 K, a n d SL = ( 1 8 4 . 4 - 7 7 . 8 ) /
80 = 1.333~ (K)/%. F r o m Eqs. (3.9)-(3.12) various c o r r e c t i o n
t e m p e r a t u r e s are calculated: ATw = -3.3~ ATu = 13.8~
A T c = 3.2~ a n d ATMe = 8.6~ F r o m Eq. (3.8) various av-
erage boiling p o i n t s are calculated: WABP = 133.7, MABP =
116.5, CABP = 127.1, a n d MeABP = 121.7~
Application a n d e s t i m a t i o n of various boiling p o i n t s are
discussed b y Van Winkle [12]. Since the m a t e r i a l s boil over a
range of t e m p e r a t u r e , any one average boiling p o i n t fails to be
useful for correlation of all properties. The m o s t useful type
of ABP is MeABP, which is r e c o m m e n d e d for correlation of
m o s t physical p r o p e r t i e s as well as calculation of Watson K
as will be discussed later in this chapter. However, for calcula-
tion of specific heat, VABP is r e c o m m e n d e d [ 12]. I n E x a m p l e
3.1, MeABP is 121.7~ w h i c h varies from 126.7 for the ASTM
D 86 t e m p e r a t u r e at 50 vol% distilled (T50). However, b a s e d on
o u r experience, for n a r r o w boiling r a n g e fractions with SL <
0.8~ the MeABP is very close to 50% ASTM t e m p e r a t u r e .
As a n example, for a gas oil s a m p l e [11] with ASTM t e m p e r a -
tures of 261.7, 270, 279.4, 289.4, a n d 307.2~ at I0, 30, 50, 70,
a n d 90 vol%, the MeABP is calculated as 279, w h i c h is very
close to 50% ASTM t e m p e r a t u r e of 279.4~ F o r this fraction
the value of SL is 0.57~ w h i c h indicates the boiling range
is quite narrow. Since n o n e of the average boiling p o i n t s de-
fined here r e p r e s e n t the true boiling p o i n t of a fraction, the
50% ASTM t e m p e r a t u r e m a y be u s e d as a c h a r a c t e r i s t i c boil-
ing p o i n t instead of average boiling point. I n this case it is
a s s u m e d that the difference b e t w e e n these t e m p e r a t u r e s is
within the range of e x p e r i m e n t a l u n c e r t a i n t y for the r e p o r t e d
distillation d a t a as well as the correlation u s e d to e s t i m a t e a
physical property.
3.2.2 Interconversion o f Various Distillation Data
W o r k to develop e m p i r i c a l m e t h o d s for converting ASTM dis-
tillations to TBP a n d EFV distillations b e g a n in the late 1920s
a n d c o n t i n u e d t h r o u g h the 1950s a n d 1960s by a large n u m b e r
of r e s e a r c h e r s [13-18]. All of the correlations were b a s e d on
d i s c o r d a n t e x p e r i m e n t a l d a t a from the literature. E x p e r i m e n -
tal ASTM, TBP, a n d EFV d a t a on w h i c h the e m p i r i c a l correla-
tions are b a s e d suffer a lack of r e p r o d u c i b i l i t y b e c a u s e there
were no s t a n d a r d i z e d p r o c e d u r e s o r a p p a r a t u s available. All
of these correlations were evaluated a n d c o m p a r e d to each
o t h e r by H o u s e et al. [ 19] to select m o s t a p p r o p r i a t e m e t h o d s
for inclusion in the API-TDB. As a result of their evaluations,
the following m e t h o d s were a d o p t e d in the API Data Book
102 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

as the best method: E d m i s t e r - P o l l o c k [14] for ASTM to TBP, TABLE 3.2---Correlation constants for Eq. (3.14).
E d m i s t e r - O k a m o t o [15-17] for ASTM to EFV, a n d Maxwell ASTM D 86
Vol% a b range,a~C
for conversion of TBP to EFV [19]. Most of these correlations
0 0.9177 1.0019 20-320
were in g r a p h i c a l forms a n d inconvenient for c o m p u t e r appli- 10 0.5564 1.0900 35-305
cations. Later, Arnold c o m p u t e r i z e d these g r a p h i c a l m e t h o d s 30 0.7617 1.0425 50-315
t h r o u g h a set of nth o r d e r p o l y n o m i a l s [20]. Correlation to 50 0.9013 1.0176 55-320
convert ASTM D 2887 (SD) to ASTM D 86 were first developed 70 0.8821 1.0226 65-330
by F o r d using m u l t i p l i e r regression analysis [21]. In the m i d 90 0.9552 1.0110 75-345
95 0.8177 1.0355 75-400
1980s Riazi a n d D a u b e r t [22] developed analytical m e t h o d s Source: Ref. [22].
for the conversion of distillation curves b a s e d on the general- aTemperatures are approximated to nearest 5.
ized correlation for h y d r o c a r b o n p r o p e r t i e s given b y Eq. (2.2).
These m e t h o d s were a d o p t e d by the API in the fifth edition of lation curve is available then ASTM curve can be e s t i m a t e d as
API-TDB-88 [2] to replace the previous methods. Continued
interests f r o m the p e t r o l e u m i n d u s t r y for these conversion (3.15) ASTM D 86 = (TBP) 1/b
m e t h o d s led to d e v e l o p m e n t of further methods. The latest
m e t h o d s for the conversion of distillation curves were devel- w h e r e constants a a n d b are given in Table 3.2 as for Eq. (3.14).
o p e d by D a u b e r t in m i d 1990s [23] t h r o u g h m o d i f y i n g R i a z i -
D a u b e r t correlations. In this section the API m e t h o d s ( R i a z i - 3.2.2.1.2 A S T M D 86 and E F V Conversions--Application
D a u b e r t a n d Daubert) for conversion of distillation d a t a are of Eq. (2.13) to this type of conversion gives
presented, w h i c h are also r e c o m m e n d e d a n d used in o t h e r
(3.16) EFV = a(ASTM D 86)b(SG) c
references a n d industrial software [24, 25].
w h e r e constants a, b, a n d c were o b t a i n e d from m o r e t h a n 300
3.2.2.1 Riazi-Daubert M e t h o d d a t a points a n d are given in Table 3.3. E q u a t i o n (3.16) was
Riazi a n d D a u b e r t m e t h o d s for the interconversion of vari- evaluated with m o r e t h a n 300 d a t a points f r o m 43 different
ous distillation d a t a are b a s e d on the generalized c o r r e l a t i o n s a m p l e s a n d gave AAD of 6~ while the m e t h o d of E d m i s t e r -
for p r o p e r t y e s t i m a t i o n of h y d r o c a r b o n s in the form of Eq. O k a m o t o [15] gave an AAD of 10~ [22, 26].
(2.38). Available distillation t e m p e r a t u r e a n d specific gravity In using these equations if specific gravity of a fraction is
of the fraction are used as the i n p u t p a r a m e t e r s to e s t i m a t e n o t available, it m a y be e s t i m a t e d from available distillation
the d e s i r e d distillation d a t a in the following form [22]: curves at 10 a n d 50% p o i n t s as given by the following equa-
tion:
(3.13) T/ (desired) = a [T/ (available)] b SG c
(3.17) SG = aTbmoT(o
w h e r e T/ (available) is the available distillation t e m p e r a t u r e
w h e r e constants a, b, a n d c for the three types of distillation
at a specific vol% distilled a n d T/(desired) is the desired distil-
data, namely, ASTM D 86, TBP, a n d EFV, are given in Table
lation d a t a for the s a m e vol% distilled, b o t h are in kelvin. SG
3.4. T e m p e r a t u r e s at 10 a n d 50% are b o t h in kelvin.
is the specific gravity of fraction at 15.5~ a n d a, b, a n d c are
correlation p a r a m e t e r s specific for each conversion type a n d
3.2.2.1.3 SD to A S T M D 86 Conversions--The equation
each vol% p o i n t on the distillation curve. F o r example, if this
derived from Eq. (3.13) for the conversion of s i m u l a t e d dis-
e q u a t i o n is u s e d to convert ASTM to EFV at 10%, T,.(available)
tillation (ASTM D 2887) to ASTM D 86 distillation curve has
is ASTM t e m p e r a t u r e at 10% a n d T/(desired) is the EFV tem-
the following form:
p e r a t u r e at 10% a n d constants a, b, a n d c are specific for this
conversion type at 10% of volume vaporized. (3.18) ASTM D 86 = a(SD)b(F) C
w h e r e c o n s t a n t F is a p a r a m e t e r specifically used for this type
3.2.2.1.1 A S T M D 86 and TBP Conversion--If distillation of conversion a n d is given by the following equation:
d a t a available are in the form of ASTM D 86 a n d d e s i r e d dis-
tillation is TBP, Eq. (3.13) can be used, b u t for this p a r t i c u l a r (3.19) F = 0.01411(SD 10~ 50~
type of conversion value of c o n s t a n t c for all p o i n t s is zero in w h i c h SD 10% a n d SD 50% are the SD t e m p e r a t u r e s in
a n d the e q u a t i o n reduces to kelvin at 10 a n d 50 wt% distilled, respectively. P a r a m e t e r F
(3.14) TBP = a(ASTM D 86) b calculated f r o m Eq. (3.19) m u s t be substituted in Eq. (3.18) to
estimate ASTM D 86 t e m p e r a t u r e at c o r r e s p o n d i n g p e r c e n t
w h e r e b o t h TBP a n d ASTM t e m p e r a t u r e s are for the s a m e p o i n t expressed in v o l u m e basis. E q u a t i o n (3.18) c a n n o t be
vol% distilled a n d are in kelvin. Constants a a n d b at various
p o i n t s along the distillation curve with the r a n g e of applica- TABLE 3.3---Correlation constants for Eq. (3.16).
tion are given in Table 3.2. ASTM D 86
F o r a total of 559 d a t a p o i n t s for 80 different samples, Eq. Vol% a b c range,a~C
(3.14) gives an average absolute deviation (AAD) of a b o u t 0 2.9747 0.8466 0.4209 10-265
10 1.4459 0.9511 0.1287 60-320
5~ while the E d m i s t e r - P o l l o c k m e t h o d [14] gives an AAD 30 0.8506 1.0315 0.0817 90-340
of a b o u t 7~ Generally p r e d i c t i o n s at 0% give higher errors 50 3.2680 0.8274 0.6214 110-355
a n d are less reliable. Details of evaluations are given in o u r 70 8.2873 0.6871 0.9340 130-400
previous p u b l i c a t i o n s [22, 26]. E q u a t i o n (3.14) can be easily 90 10.6266 0.6529 1.1025 160-520
100 7.9952 0.6949 1.0737 190-430
reversed to p r e d i c t ASTM f r o m TBP data, b u t this is a rare Source: Ref. [22].
a p p l i c a t i o n as usually ASTM d a t a are available. If TBP distil- aTemperatures are approximated to nearest 5.
 3. C H A R A C T E R I Z A T I O N O F P E T R O L E U M F R A C T I O N S 103

TABLE 3.4---Correlation constants for Eq. (3.17).

Distillation Tl0 7"50 SG No. of AAD
type range,a~ range,a~ range a b c data points %
ASTM D 86 35-295 60-365 0.70-1.00 0.08342 0.10731 0.26288 120 2.2
TBP 10-295 55-320 0.67-0.97 0.10431 0.12550 0.20862 83 2.6
EFV 79-350 105-365 0.74--0.91 0.09138 -0.0153 0.36844 57 57
Source: Ref. [22].
aTemperaturesare approximatedto nearest 5.

used in a reverse form to predict SD from ASTM D 86, b u t this calculated ASTM a n d SG, the EFV t e m p e r a t u r e s can be es-
type of conversion is usually n o t desired as most predictive timated from Eq. (3.16) with constants given in Table 3.2.
m e t h o d s use ASTM D 86 data while laboratories report SD EFV = 2.9747 (32 + 273) 0.8466 (0.7862)~176 : 340.9 K ---
data. Constants a, b, a n d c in Eq. (3.18) were obtained from 340.9 - 273 = 67.9~ The calculated value is very close to the
81 different samples a n d 567 data points a n d are given i n actual value of 68.3~ (see Table 3.5). Similarly EFV values at
Table 3.5 with the range of SD data at each percentage along other points are calculated a n d results are s h o w n i n Fig. 3.19.
the distillation curve. Predicted EFV curve from TBP are very close to the actual
E q u a t i o n (3.18) a n d the m e t h o d of Ford published by EFV curve. The AAD between predicted EFV a n d experimen-
ASTM, included in the earlier editions of API-TDB [21], were tal data is 2.6 K. It should be n o t e d that if experimental ASTM
evaluated by some 570 data points a n d gave AAD of 5 a n d data a n d specific gravity were used, the predicted values of
5.5~ respectively [22, 26]. Larger errors were observed at EFV would be even closer to the experimental values. #
the initial a n d final boiling points (0 a n d 100%) b u t excluding
these points the AAD reduces to a b o u t 3~ for conversions 3.2.2.2 Daubert's M e t h o d
within the range of 10-90% distilled. Daubert a n d his group developed a different set of equations
The procedures given in this section should be used with the to convert ASTM to TBP, SD to ASTM, a n d SD to TBP [23].
range of data specified in Tables 3.1-3.4. Use of these equa- These m e t h o d s have b e e n included in the sixth edition of API-
tions outside the specified ranges could cause large errors. TDB [2] a n d are given in this section. I n these methods, first
Graphical forms of these equations for conversion of various conversion should be m a d e at 50% p o i n t a n d then the differ-
distillation curves are given in Reference [22] as well as in the ence between two cut points are correlated i n a form similar
fourth edition of the API-TDB-88 [2]. One of the advantages of to Eq. (3.14). I n this m e t h o d SD data can be converted di-
these equations is that they can be used i n reversed form. This rectly to TBP without calculating ASTM as was needed i n the
m e a n s one m a y estimate EFV from TBP data t h r o u g h conver- R i a z i - D a u b e r t method.
sion of TBP to ASTM by Eq. (3.15) a n d t h e n u s i n g Eq. (3.16)
to estimate EFV from calculated ASTM curve. The example 3.2.2.2.1 A S T M a n d T B P C o n v e r s i o n - - T h e following
below shows this conversion process. e q u a t i o n is used to convert a n ASTM D 86 distillation at 50%
p o i n t t e m p e r a t u r e to a TBP distillation 50% p o i n t tempera-
E x a m p l e 3 . 2 - - F o r a b l e n d of n a p h t h a - k e r o s e n e sample, ture.
ASTM, TBP, a n d EFV distillation curves are given in the API-
TDB [2]. These data are represented in Table 3.6. Use the TBP(50%) = 255.4 + 0.8851[ASTM D 86(50%) - 255.4] 1~
R i a z i - D a u b e r t m e t h o d s to predict EFV curve from TBP curve. (3.20)
where ASTM (50%) a n d TBP (50%) are t e m p e r a t u r e s at 50%
$ o l u t i o n - - T B P data are used as available i n p u t data. Equa-
volume distilled in kelvin. E q u a t i o n (3.20) c a n also be used
tion (3.15) should be used to estimate ASTM D 86 from TBP.
i n a reverse form to estimate ASTM from TBP. The following
For the initial p o i n t at 0%, the calculations are as follows.
e q u a t i o n is used to d e t e r m i n e the difference between two cut
ASTM D 86 = (1/0.9177) 1/1~176(10 + 273) 1/1~176= 305 K =
points:
305 - 273 = 32~ The actual data for the initial ASTM tem-
perature is 35~ which is close to the calculated value. (3.21) Y / = AX/B
Now to estimate EFV from Eq. (3.16), specific gravity, is re-
where
quired which is not given by the problem. SG can be esti-
difference in TBP t e m p e r a t u r e b e t w e e n two cut
m a t e d from Eq. (3.17) a n d constants given in Table 3.3 for Y/=
points, K (or ~
the TBP. F r o m Table 3.6, T10(TBP) = 71.1 a n d Ts0(TBP) =
observed difference in ASTM D 86 t e m p e r a t u r e be-
204.4~ Using these values i n Eq. (3.17) gives SG = 0.10431 Xi --
tween two cut points, K (or ~
(71.1 + 273) 0"1255 (204.4 + 273)020862= 0.7862. Now from
A , B = constants varying for each cut p o i n t a n d are given
in Table 3.7
TABLE 3.5---Correlation constants for Eq. (3.18).
SD
Vol% a b c rangefl~C TABLE 3.6---Data on various distillation curves
for a naphtha-kerosene blend [2].
0 5.1764 0.7445 0.2879 -20-200
10 3.7452 0.7944 0.2671 25-230 Vol% ASTMD 86, TBP, EFV,
distilled ~C ~C ~C
30 4.2749 0.7719 0.3450 35-255
50 1 . 8 4 4 5 0.5425 0.7132 55-285 0 35.0 10.0 68.3
70 1 . 0 7 5 1 0.9867 0.0486 65-305 10 79.4 71.1 107.2
90 1 . 0 8 4 9 0.9834 0.0354 80-345 30 145.6 143.3 151.1
100 1.799l 0.9007 0.0625 95--405 50 201.7 204.4 182.2
Source:Ref.[22]. 70 235.6 250.6 207.2
~Ternperaturesare approximatedto nearest 5. 90 270.6 291.7 228.3
104 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

300 (3.22). The s u m m a r y of results is given in Table 3.8. The

~
~176 overall average absolute deviations (AAD) for Eqs. (3.14) a n d
....... T B P (exp) . . *"" 9 (3.20) are calculated as 2.2 a n d 3.8~ respectively. As it is
E F V (exp) ~176 ~
/
O seen in Table 3,8, Eqs. (3.20)-(3.22) are m o r e a c c u r a t e at
~ 20o
o E F V (calc) -"" ~ 30, 50, a n d 70% p o i n t s t h a n at the lower or h i g h e r ends of
the distillation curve. 0

3.2.2.2.2 SD to TBP C o n v e r s i o n - - A s d e s c r i b e d before,

s i m u l a t e d distillation by gas c h r o m a t o g r a p h y (ASTM D 2887)
[~ 100 is n o w c o m m o n l y u s e d as a m e a n s of m e a s u r i n g boiling p o i n t s
of light p e t r o l e u m fractions. SD curves are expressed in t e r m s
~176 of t e m p e r a t u r e versus wt% distilled, while TBP curves are ex-
p r e s s e d in t e r m s of t e m p e r a t u r e versus vol% distilled. In the
0 i i i i Daubert's m e t h o d of conversion of SD to TBP it is a s s u m e d
0 20 40 60 80 100 t h a t TBP at 50 vol% distilled is equal to SD t e m p e r a t u r e at
50 wt% distilled. E q u a t i o n s for conversion of SD to TBP are
Vol% Distilled s i m i l a r to equations developed for conversion of ASTM to
FIG. 3.19~Prediction of EFV from TBP curve TBE
for a naphtha-kerosene blend (Example 3.2). (3.23) TBP(50 vol%) = SD(50 wt%)
w h e r e SD (50 wt%) a n d TBP (50 vol%) are t e m p e r a t u r e s at
To d e t e r m i n e the true boiling p o i n t t e m p e r a t u r e at a n y
50% distilled in kelvin (or ~ The difference b e t w e e n adja-
p e r c e n t distilled, calculation should b e g i n with 50% TBP
cent cut points is calculated f r o m the following e q u a t i o n as
t e m p e r a t u r e a n d a d d i t i o n or s u b t r a c t i o n of the p r o p e r tem-
given by the API-TDB [2].
p e r a t u r e difference Y/.
(3.24) Vii = CWin
TBP (0%) = TBP (50%) - Y4 - Y5 - I16
TBP (10%) = TBP (50%) - Y4 - Y5 where
V / = difference in TBP t e m p e r a t u r e b e t w e e n two cut
TBP (30%) - TBP (50%) - Y4 points, K (or ~
(3.22)
TBP (70%) = TBP (50%) + Y3 W / = observed difference in SD t e m p e r a t u r e b e t w e e n
TBP (90%) = TBP (50%) + Y3 + Y2 two cut points, K (or ~
C, D = constants varying for each cut p o i n t a n d are given
TBP (100%) = TBP (50%) + Y3 + Y2 + Y1
in Table 3.9
This m e t h o d was developed b a s e d on s a m p l e s with ASTM To d e t e r m i n e the true boiling p o i n t t e m p e r a t u r e at a n y per-
50% p o i n t t e m p e r a t u r e of less t h a n 250~ (480~ b u t it is cent distilled, calculation s h o u l d begin w i t h 50% TBP temper-
r e c o m m e n d e d for e x t r a p o l a t i o n up to fractions with ASTM ature a n d a d d i t i o n o r s u b t r a c t i o n of the p r o p e r t e m p e r a t u r e
50% t e m p e r a t u r e of 315~ (600~ as suggested b y the API [2]. difference V/.
Average absolute deviation for this m e t h o d as r e p o r t e d by the
TBP(5%) = TBP(50%) - Vs - V6 - - V7
API-TDB [2] is a b o u t 4.6~ for s o m e 70 samples. Predicted
TBP at 0 a n d 100% are the least a c c u r a t e values followed b y TBP(10%) = TBP(50%) - Vs - V6
values at i 0 a n d 90% p o i n t s as it is s h o w n in the following TBP(30%) = TBP(50%) - Vs
example. (3.25) TBP(70%) = TBP(50%) + V4
TBP(90%) = TBP(50%) + V4 + 173
Example 3 . 3 - - A S T M D 86 a n d TBP distillation d a t a for a
kerosene s a m p l e [1] are given in Table 3.7. Predict the TBP TBP(95%) = TBP(50%) + V4 + V3 + V2
curve from ASTM d a t a using R i a z i - D a u b e r t a n d Daubert's TBP(100%) = TBP(50%) + V4 + V3 + V2 + V1
m e t h o d s a n d calculate AAD for each m e t h o d . This m e t h o d is a p p l i c a b l e to fractions with TBP 50% p o i n t s
in the range of 120-370~ (250-700~ Average absolute de-
Solution--The R i a z i - D a u b e r t m e t h o d for conversion of viation for this m e t h o d as r e p o r t e d by the API-TDB [2] is
ASTM to TBP d a t a is p r e s e n t e d by Eq. (3.14) a n d constants in a b o u t 7.5~ for a b o u t 21 samples. B a s e d on 19 datasets it
Table 3.2. The Daubert's m e t h o d is expressed by Eqs. (30.20)- was observed t h a t errors in direct conversion of SD to TBP is
slightly h i g h e r t h a n if SD is converted first to ASTM a n d t h e n
TABLE 3.7--Correlation constants for Eq. (3.21). e s t i m a t e d ASTM is converted to TBP by Eqs. (3.20)-(3.22).
Cut point Maximum allowable Details of these evaluations are given by the API [2]. Predicted
range, % A B xi,a~
1 100-90 0.1403 1.6606 TBP at 5, 95, a n d 100% are the least a c c u r a t e values followed
2 90-70 2.6339 0.7550 "55-- by values at 10 a n d 90% p o i n t s as is s h o w n in the following
3 70-50 2.2744 0.8200 85 example.
4 50-30 2.6956 0.8008 140
5 30-10 4.1481 0.7164 140 Example 3 . 4 - - E x p e r i m e n t a l ASTM D 2887 (SD) a n d TBP dis-
6 10-0 5.8589 0.6024 55
Source: Refs. [2, 23]. tillation d a t a for a p e t r o l e u m fraction are given in Table 3.9 as
aTemperatures are approximated to nearest 5. t a k e n f r o m API [2]. Predict the TBP curve from SD d a t a using
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S 105

TABLE 3.8--Prediction of TBP from ASTM for a kerosene sample of Example 3.3.
Vol% ASTM D 86 TBP Eq. (3.14) Eqs. (3.20)-(3.22)
distilled exp,~ exp,~ TBP calc,~ AD,~ TBP calc,~ AD,~
0 165.6 146.1 134.1 12.0 133.1 13.0
10 176.7 160.6 160.6 0.0 158.1 2.5
30 193.3 188.3 188.2 0.1 189.2 0.9
50 206.7 209.4 208.9 0.5 210.6 1.2
70 222.8 230.6 230.2 0.4 232.9 2.3
90 242.8 255.0 254.7 0.3 258.1 3.1
Overall AAD,~ 2.2 3.8

R i a z i - D a u b e r t a n d Daubert's m e t h o d s a n d calculate AAD for d a t a are similar to the equations developed b y D a u b e r t [2, 23]
each method. to convert ASTM o r SD to TBP as given in this section. The
equations are s u m m a r i z e d as following:
Solution--The R i a z i - D a u b e r t m e t h o d s do n o t provide a di-
rect conversion f r o m SD to TBP, but one c a n use Eqs. (3.18) ASTM D 86(50vo1%) = 255.4 + 0.79424
a n d (3.19) to convert SD to ASTM D 86 a n d t h e n Eq. (3.14) (3.26) [SD(50 wt%) - 255.4] 1.0395
s h o u l d be used to convert ASTM to TBP data. F r o m Eq. (3.19)
a n d use of SD at 10% a n d 50% points, the value of p a r a m - w h e r e SD (50 wt%) a n d ASTM D 86 (50 vo1%) a r e t e m p e r a -
eter F is calculated as 0.8287. Value of SD t e m p e r a t u r e at tures at 50% distilled in kelvin. The difference b e t w e e n adja-
50 wt% is 168.9~ from Eq. (3.18) with a p p r o p r i a t e c o n s t a n t s cent cut points is calculated from the following e q u a t i o n as
in Table 3.5 one c a n o b t a i n ASTM D 86 (50%) = 166.3~ given b y the API-TDB [2].
S u b s t i t u t i n g this value for ASTM into Eq. (3.14) gives TBP
(50 vol%) = 167.7~ while the e x p e r i m e n t a l value as given (3.27) Ui = E T f
in Table 3.10 is 166.7~ The AD is t h e n calculated as
where
167.7-166.7 = 1~C. Daubert's m e t h o d for conversion of SD Ui = difference in ASTM D 86 t e m p e r a t u r e s between
to TBP is direct a n d is p r e s e n t e d t h r o u g h Eqs. (3.23)-(3.25).
two cut points, K (or ~
According to Eq. (3.23), TBP (50%) = SD (50%) = 168.9~
T / = observed difference in SD t e m p e r a t u r e s between
w h i c h gives a n AD of 2.2~ for this point. A s u m m a r y of com-
two cut points, K (or ~
plete calculation results is given in Table 3.10. The overall AAD
E, F = constants varying for each cut p o i n t a n d are given
for Eqs. (3.14) a n d (3.18) is 4.8, while for Eqs. (3.23)-(3.25)
in Table 3.11
is 2.2~
To d e t e r m i n e the ASTM D 86 t e m p e r a t u r e at any p e r c e n t
distilled, calculations should begin with 50% ASTM D 86 tem-
Results p r e s e n t e d in E x a m p l e 3.4 show t h a t Eqs. (3.23)- p e r a t u r e a n d a d d i t i o n o r s u b t r a c t i o n of the p r o p e r t e m p e r a -
(3.25) are m o r e a c c u r a t e t h a n Eqs. (3.14) a n d (3.18) for the ture difference Ui.
conversion of SD to TBE One of the reasons for such a result
is t h a t the s a m p l e p r e s e n t e d in Table 3.10 to evaluate these ASTM D 86(0%) = ASTM D 86(50%)
m e t h o d s is t a k e n from the s a m e d a t a b a n k used to develop
- U4 - Us - U6
correlations of Eqs. (3.23)-(3.25). In a d d i t i o n these equations
provide a direct conversion of SD to TBE However, one should ASTM D 86(10%) -- ASTM D 86(50%)
realize t h a t Eqs. (3.23)-(3.25) are b a s e d on only 19 datasets - u4 - u5
a n d this limits the a p p l i c a t i o n of these equations. While Eqs.
ASTM D 86(30%) -- ASTM D 86(50%) - U4
(3.14), (3.18), a n d (3.19) are b a s e d on m u c h larger d a t a b a n k s (3.28)
with w i d e r range of application. As available d a t a on b o t h ASTM D 86(70%) = ASTM D 86(50%) + U3
SD a n d TBP are very limited, a concrete r e c o m m e n d a t i o n on ASTM D 86(90%) = ASTM D 86(50%)
s u p e r i o r i t y of these two a p p r o a c h e s c a n n o t be m a d e at this
.-~ U3 -~ U 2
time.
ASTM D 86(100%) = ASTM D 86(50%)
3,2.2.2.3 SD to A S T M D 86 Conversion--Equations to + U3 + U2 + Ul
convert SD (ASTM D 2887) distillation d a t a to ASTM D 86
This m e t h o d is a p p l i c a b l e to fractions with ASTM D 86 50%
TABLE 3.9--Correlation constants for Eq. (3.24). points in the r a n g e o f 65-315~ (150-600~ The average ab-
Cutpoint M a x i m u m allowable solute deviation for this m e t h o d as r e p o r t e d b y the API-TDB
i range, % C D W/,~ [2] is a b o u t 6~ for s o m e 125 s a m p l e s a n d a p p r o x i m a t e l y 850
1 100-95 0.03849 1.9733 15 d a t a points. P r e d i c t e d ASTM t e m p e r a t u r e s at 0 a n d 100% are
2 95-90 0.90427 0.8723 20 the least a c c u r a t e values followed by values at 10 a n d 90%
3 90-70 0.37475 1.2938 40
4 70-50 0.25088 1.3975 40 points as is s h o w n in the following example.
5 50-30 0.08055 1.6988 40
6 30-10 0.02175 2.0253 40
7 10-0 0.20312 1.4296 20 Example 3 . 5 - - E x p e r i m e n t a l ASTM D 2887 (SD) a n d ASTM
Source: Taken with permission from Refs. [2, 23]. D 86 distillation d a t a for a p e t r o l e u m fraction are given in
aTemperatures are approximated to nearest 5. Table 3.11 as t a k e n f r o m the API-TDB [2]. Predict the ASTM
106 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

TABLE 3.10~Prediction of TBP from SD for a petroleum fraction of Example 3.4,

Wt% or vol% ASTMD 2887 TBP Eqs. (3.18) and (3.14) Eq. (3.23)-(3.25)
distilled (SD) exp,~ exp,~ TBP calc,~ AD,~ TBP calc,~ AD,~
10 151.7 161.1 146.1 15.0 164.3 3.2
30 162.2 163.3 157.1 6.2 166.9 3.5
50 168.9 166.7 167.7 1.0 168.9 2.2
70 173.3 169.4 170.7 1.3 170.9 1.5
90 181.7 173.9 179.1 5.3 176.7 2.8
Overall AAD, ~ 4.8 2.2

D 86 curve f r o m SD d a t a using R i a z i - D a u b e r t a n d Daubert's earlier in this c h a p t e r u n d e r ASTM D 1160 test m e t h o d . F o r

m e t h o d s a n d calculate AAD for each m e t h o d . p r e d i c t i o n of physical a n d t h e r m o d y n a m i c p r o p e r t i e s n o r m a l
boiling p o i n t s are required. F o r this r e a s o n m e t h o d s of cal-
Solution--Both m e t h o d s provide direct m e t h o d s for conver- culation of equivalent a t m o s p h e r i c boiling p o i n t (EABP) are
sion of SD to ASTM D 86 a n d calculations a r e s i m i l a r to those i m p o r t a n t . One has to recognize t h a t EABP is n o t a real boil-
p r e s e n t e d in E x a m p l e s 3.3. a n d 3.4. E q u a t i o n s (3.26)-(3.28) ing p o i n t as for such heavy fractions there is no actual a n d real
are used for Daubert's m e t h o d , while Eqs. (3.18) a n d (3.19) e x p e r i m e n t a l value for the n o r m a l boiling point. This p a r a m -
are used for R i a z i - D a u b e r t m e t h o d . A s u m m a r y of c o m p l e t e eter can be o b t a i n e d f r o m conversion of distillation curves at
calculation results is given in Table 3.12. The overall AAD for low p r e s s u r e s to equivalent distillation curves at a t m o s p h e r i c
Eq. (3.18) is 1.5, while for Eqs. (3.26)-(3.28) is 2.0~ # p r e s s u r e s a n d it is j u s t an a p p a r e n t n o r m a l boiling point. The
basis of such conversion is v a p o r p r e s s u r e correlation for the
Results p r e s e n t e d in E x a m p l e 3.5 s h o w t h a t Eq. (3.18) is fraction of interest, w h i c h will be discussed in C h a p t e r 6. In
slightly m o r e a c c u r a t e t h a n Eqs. (3.26)-(3.28) for the conver- this p a r t we p r e s e n t calculation m e t h o d s for the conversion
sion of SD to ASTM D 86. This is consistent with AAD r e p o r t e d of ASTM D 1160 to a t m o s p h e r i c distillation curve a n d for the
for these m e t h o d s . However, Eqs. (3.26)-(3.28) are b a s e d on p r e d i c t i o n of a t m o s p h e r i c TBP curves from ASTM D 1160.
a larger d a t a set than is Eq. (3.18). In general, R i a z i - D a u b e r t It s h o u l d be n o t e d t h a t ASTM D 1160 does not refer to a n y
m e t h o d s are s i m p l e r a n d easily reversible, while the existing specific pressure. The p r e s s u r e m a y vary f r o m 1 to 50 m m Hg.
API m e t h o d s are slightly m o r e complex. The a d v a n t a g e of W h e n D 1160 curve is converted to a distillation curve at at-
Daubert's m e t h o d s is that the p r e d i c t e d curve is s m o o t h a n d m o s p h e r i c p r e s s u r e t h r o u g h a v a p o r p r e s s u r e c o r r e l a t i o n the
uniform, while in the R i a z i - D a u b e r t m e t h o d s every p o i n t is resulting distillation curve is not equivalent to ASTM D 86 o r
p r e d i c t e d i n d e p e n d e n t of a d j a c e n t p o i n t a n d lack of s m o o t h - to TBP curve. The resulting distillation curve is referred to
ness in p r e d i c t e d curve is possible, a l t h o u g h this is rarely ob- as equivalent a t m o s p h e r i c ASTM D 1160. A n o t h e r low pres-
served in o u r experience. Since in the Daubert's m e t h o d s tem- sure distillation d a t a is TBP distillation curve at 1, 10, o r
p e r a t u r e s at 0 a n d 10% points are calculated from p r e d i c t e d 50 m m Hg. T h r o u g h v a p o r p r e s s u r e correlations TBP at re-
values at 30 a n d 50% points, larger errors are observed at the d u c e d p r e s s u r e s can be converted to a t m o s p h e r i c TBP. There
lower (0, 5, a n d 10% distilled) o r u p p e r ends (90, 95, a n d 100% is a p r o c e d u r e for the conversion of ASTM D 1160 to TBP
distilled) of p r e d i c t e d distillation curves. In general the accu- at 10 m m Hg which is p r e s e n t e d in this section. Therefore,
r a c y of b o t h m e t h o d s in the p r e d i c t i o n of distillation curves to convert ASTM D 1160 to TBP at a t m o s p h e r i c p r e s s u r e
at 0 a n d 100% points are limited. This is m a i n l y due to the one has to convert D 1160 at any p r e s s u r e to D 1160 at 10
e x p e r i m e n t a l u n c e r t a i n t y in m e a s u r e m e n t of t e m p e r a t u r e s at m m H g a n d t h e n to convert resulting D 1160 to TBP at 10
the end points. m m Hg. This m e a n s if ASTM D 1160 at 1 m m Hg is avail-
able, it m u s t be first converted to D 1160 at 760 m m H g ,
3.2.2.3 Interconverion o f DistiUation Curves t h e n to D 1160 at 10 m m Hg followed by conversion to TBP
at Reduced Pressures at 10 m m H g a n d finally to TBP at 760 m m H g . A s u m m a r y
c h a r t for various conversions is p r e s e n t e d at the end of this
N o r m a l boiling points of heavy p e t r o l e u m fractions such as
p r o d u c t s of a v a c u u m distillation c o l u m n c a n n o t be m e a s u r e d section.
due to the t h e r m a l d e c o m p o s i t i o n of heavy h y d r o c a r b o n s at
high t e m p e r a t u r e s . F o r this r e a s o n distillation d a t a a r e re- 3,2.2.3.1 Conversion o f a Boiling Point at Sub- or Super-
p o r t e d at r e d u c e d pressures of 1-50 m m H g , as d e s c r i b e d Atmospheric Pressures to the N o r m a l Boiling Point or
Vice Versa--The conversion of boiling p o i n t or s a t u r a t i o n
t e m p e r a t u r e at s u b a t m o s p h e r i c (P < 760 m m Hg) or super-
TABLE 3.11---Correlation constants for Eq. (3.27). a t m o s p h e r i c (P > 760 m m Hg) conditions to n o r m a l boiling
Cut point Maximum allowable p o i n t is b a s e d on a v a p o r p r e s s u r e correlation. The m e t h o d
i range, % E F T/,~ widely u s e d in the i n d u s t r y is the correlation developed for
1 100-90 2.13092 0.6596 55 p e t r o l e u m fractions by Maxwell a n d Bonnell [27], w h i c h is
2 90-70 0.35326 1,2341 55 also u s e d by the API-TDB [2] a n d o t h e r sources [24] a n d is
3 70-50 0.19121 1.4287 55
4 50-30 0.10949 1.5386 55 p r e s e n t e d here. This correlation is given for several p r e s s u r e
5 30-10 0.08227 1.5176 85 ranges as follows:
6 10-0 0.32810 1.1259 85
Source: Taken with permission from Ref. [2]. 748.1 QT
aTemperatures are approximated to nearest 5. (3.29) 2r~ = 1 + T(0.3861Q - 0.00051606)
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S 107

TABLE 3.12--Prediction of ASTM D 86 from SD for a petroleum fraction of Example 3.5.

Eqs. (3.18) and (3.19) Eqs. (3.25)-(3.28)
Vol% ASTMD 2887 ASTMD 86 ASTM D 86 ASTMD 86
distilled (SD) exp,~ exp,~ calc,~ AD,~ calc,~ AD,~
10 33.9 56.7 53.2 3.4 53.5 3.2
30 64.4 72.8 70.9 1.9 68.2 4.5
50 101.7 97.8 96.0 1.8 96.8 1.0
70 140.6 131.7 131.3 0.4 132.5 0.9
90 182.2 168.3 168.3 0.0 167.8 0.6
Overall AAD, ~ 1.5 2.0

Q = 6.761560 - 0.987672 lOglo P (P < 2 m m H g ) where

3000.538 - 43 loglo P P
7"~ = Tb -- 1.3889 F (Kw - 12) logl0 760
Q = 5.994296 - 0.972546 loglo P (2 < P < 760 m m Hg)
2663.129 - 95.76 loglo P w h e r e all the p a r a m e t e r s are defined in Eq. (3.29). The m a i n
a p p l i c a t i o n of this e q u a t i o n is to e s t i m a t e boiling p o i n t s at
Q = 6.412631 - 0.989679 loglo P (P > 7 6 0 m m H g ) I0 m m H g from a t m o s p h e r i c boiling points. At P = 10
2770.085 - 36 loglo P m m H g , Q = 0.001956 a n d as a result Eq. (3.30) reduces to
the following simple form:
P
Tb = T,'b + 1.3889F(Kw - 12)logao 760 0.683398T~
(3.31) T ( 1 0 m m H g ) = 1 - 1.63434 x I0-4T~
F=0 (Tu < 367 K) o r w h e n Kw
in w h i c h T~ is calculated from Tb as given in Eq. (3.30) a n d
is not available b o t h are in kelvin. T e m p e r a t u r e T (10 m m Hg) is the boiling
F = - 3 . 2 9 8 5 + 0.009 Tb (367 K _< Tb < 478 K) p o i n t at r e d u c e d p r e s s u r e of 10 m m Hg in kelvin. By a s s u m -
F = - 3 . 2 9 8 5 + 0.009 Tu (Tb > 478 K) ing Kw = 12 (or F = 0) a n d for low-boiling fractions value
of n o r m a l boiling point, Tb, can be used i n s t e a d of T~ in Eq.
(3.31). To use these equations for the conversion of boiling
where
p o i n t from one low p r e s s u r e to a n o t h e r low p r e s s u r e (i.e.,
P = p r e s s u r e at w h i c h boiling p o i n t o r distillation d a t a
from 1 to 10 m m Hg), two steps are required. In the first step,
is available, m m H g
n o r m a l boiling p o i n t o r T (760 m m Hg) is calculated from
T = boiling p o i n t originally available at p r e s s u r e P, in
T (1 m m H g ) by Eq. (3.29) a n d in the s e c o n d step T (10
kelvin
m m Hg) is calculated f r o m T (760 m m Hg) o r Tb t h r o u g h Eqs.
T~ = n o r m a l boiling p o i n t c o r r e c t e d to Kw = 12, in
(3.30) a n d (3.31).
kelvin
In the m i d 1950s, a n o t h e r graphical correlations for the
Tb = n o r m a l boiling point, in kelvin
e s t i m a t i o n of v a p o r p r e s s u r e of high boiling h y d r o c a r b o n s
Kw = Watson (UOP) c h a r a c t e r i z a t i o n factor [ = (1.8Tb) 1/3
were p r o p o s e d b y Myers a n d Fenske [28]. L a t e r two simple
/SG]
l i n e a r relations were derived from these charts to e s t i m a t e
F = c o r r e c t i o n factor for the fractions w i t h Kw different
T ( I 0 m m Hg) from the n o r m a l boiling p o i n t (Tb) o r boiling
from I2
p o i n t at 1 m m Hg as follows [29]:
logl0 = c o m m o n l o g a r i t h m (base 10)
T(10 mm Hg) = 0.8547T(760 rnm Hg) - 57.7 500 K < T(760 mm) < 800K
The original evaluation of this equation is on p r e d i c t i o n of va-
T(10mmHg) = 1.07T(1 rnmHg) + 19 300K < T(1 ram) < 600K
p o r p r e s s u r e of p u r e h y d r o c a r b o n s . Reliability of this m e t h o d
for n o r m a l boiling p o i n t of p e t r o l e u m fractions is unknown. (3.32)
W h e n this e q u a t i o n is a p p l i e d to p e t r o l e u m fractions, gener- w h e r e all t e m p e r a t u r e s are in kelvin. These equations repro-
ally Kw is not known. F o r these situations, T~ is calculated duce the original figures within 1%; however, they s h o u l d
with the a s s u m p t i o n that Kw is 12 a n d Tb = T~. This is to be used within the t e m p e r a t u r e s ranges specified. Equa-
equivalent to the a s s u m p t i o n of F = 0 for low-boiling-point tions (3.30) a n d (3.31) are m o r e a c c u r a t e t h a n Eq. (3.32) b u t
c o m p o u n d s o r fractions. To i m p r o v e the result a s e c o n d r o u n d for quick h a n d estimates the latter is m o r e convenient. An-
of calculations can be m a d e with Kw calculated from esti- o t h e r simple relation for quick conversion of boiling p o i n t at
m a t e d value of T~. W h e n this e q u a t i o n is a p p l i e d to distilla- various pressures is t h r o u g h the following correction, w h i c h
tion curves of c r u d e oils it s h o u l d be realized that value of was p r o p o s e d by Van K r a n e n a n d Van Nes, as given b y Van
Kw m a y change along the distillation curve as b o t h Tb a n d Nes a n d Van Westen [30].
specific gravity change.
Tb - 41 1393 - T
E q u a t i o n (3.29) can be easily used in its reverse form to log10 PT = 3.2041 1 - 0.998 • ~ • 1393 - Tb]
calculate boiling p o i n t s (T) at low o r elevated p r e s s u r e s from
n o r m a l boiling p o i n t (Tb) as follows: (3.33)
where T is the boiling p o i n t at p r e s s u r e Pr a n d Tb is the n o r m a l
(3.30) T = boiling point. Pr is in b a r a n d T a n d Tb are in K. Accuracy of
748.1 O - T[(0.3861 O - 0.00051606) this e q u a t i o n is a b o u t I%.
108 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
3.2.2.3.2 Conversion of a Distillation Curve from Sub- or
Super- Atmospheric Pressures to a Distillation Curve at
Atmospheric Pressure--The m e t h o d of conversion of boil-
ing p o i n t s t h r o u g h Eqs. (3.29)-(3.32) c a n b e used to every
p o i n t on a distillation curve u n d e r either sub- o r s u p e r a t m o -
spheric p r e s s u r e conditions. I n these equations Tb o r T (760
m m H g ) r e p r e s e n t a p o i n t along the distillation curve at at-
m o s p h e r i c pressure. It can be a p p l i e d to any of TBP, EFV,
o r ASTM D 1160 distillation curves. However, it s h o u l d be
n o t e d t h a t these equations convert distillation curves from
one p r e s s u r e to a n o t h e r w i t h i n the s a m e type. F o r example, it
is n o t possible to use these equations to directly convert ASTM
D 1160 at 10 m m H g to TBP at 760 m m H g . S u c h conversions
require two steps that are discussed in the following section.
The only distillation curve type t h a t m i g h t be r e p o r t e d un-
d e r s u p e r a t m o s p h e r i c p r e s s u r e (P > 1.01325 bar) c o n d i t i o n
is the EFV distillation curve. TBP curve m a y be at I, 10, 100,
o r 760 m m Hg pressure. E x p e r i m e n t a l d a t a on ASTM D 1160
are usually r e p o r t e d at i, 10, o r 50 m m Hg. ASTM D 86 distil-
l a t i o n is always r e p o r t e d at a t m o s p h e r i c pressure. It should
be n o t e d t h a t w h e n ASTM D 1160 distillation curve is con-
verted to o r r e p o r t e d at a t m o s p h e r i c p r e s s u r e (760 m m Hg)
it is not equivalent to or the s a m e as ASTM D 86 distillation
data. They are different types of distillation curves a n d there
is no direct conversion b e t w e e n these two curves.
3.2.2.3.3 Conversion of ASTM D 1160 at 10 m m H g to
TBP Distillation Curve at 10 m m H g - - T h e only m e t h o d
widely used u n d e r s u b a t m o s p h e r i c p r e s s u r e c o n d i t i o n for
conversion o f distillation curves is the one developed b y
E d m i s t e r - O k a m o t o [17], w h i c h is used to convert ASTM
D 1160 to TBP, b o t h at 10 m m H g . This m e t h o d is graphi-
cal a n d it is also r e c o m m e n d e d by the API-DTB [2]. In this
m e t h o d it is a s s u m e d the at 50% points ASTM D 1160 a n d
TBP t e m p e r a t u r e s are equal. The E d m i s t e r - O k a m o t o c h a r t
is converted into e q u a t i o n form t h r o u g h regression of values
r e a d from the figure in the following form [2]:
TBP(100%) = ASTM D 1160(100%)
TBP(90%) = ASTM D 1160(90%)
TBP(70%) = ASTM D 1160(70%)
(3.34) TBP(50%) = ASTM D 1160(50%)
TBP(30%) = ASTM D 1160(50%) - F1
TBP(10%) = ASTM D 1160(30%) - F2
TBP(0%) = ASTM D 1160(10%) - F3
w h e r e functions F1, F2, a n d / : 3 are given in t e r m s of temper-
a t u r e difference in the ASTM D 1160:
F1 = 0.3 + 1.2775(AT1) - 5.539 • 10-3(AT1) 2 @ 2.7486
X 10-5(AT1) 3
F2 = 0.3 + 1.2775(AT2) - 5.539 x 10-3(AT2) 2 + 2.7486
• 10-5(AT2) 3
F3 = 2.2566(AT3) - 266.2 x 10-4(AT3) 2 + 1.4093
• 10-4(AT3) 3
AT1 = ASTM D 1160(50%) - ASTM D 1160(30%)
AT2 = ASTM D 1160(30%) - ASTM D 1160(10%)
AT3 = ASTM D 1160(10%) - ASTM D 1160(0%)
in the above relations all t e m p e r a t u r e s are either in ~ o r in
kelvin.
3.2.2.4 Summary Chart for Interconverion of Various
Distillation Curves
A s u m m a r y of all conversion m e t h o d s is s h o w n in Fig. 3.20.
It s h o u l d be n o t e d that any distillation curve at low p r e s s u r e
(i.e., ASTM D 1160 o r EFV at 1, 10, 50, m m H g o r TBP at
1 m m Hg) should be first converted to TBP distillation curve
at 10 m m Hg before they are converted to TBP at a t m o s p h e r i c
pressure.
Example 3 . 6 - - F o r a p e t r o l e u m fraction the ASTM D 1160
distillation d a t a at 10 m m Hg are given in Table 3.13. P r e d i c t
the TBP curve at a t m o s p h e r i c pressure.
Solution--ASTM D 1160 d a t a have b e e n converted to TBP
at 10 m m H g by Eq. (3.34). Then Eq. (3.29) with P - - 10
m m H g a n d Q = 0.001956 is u s e d to convert TBP from 10
to 760 m m H g . A s u m m a r y of results is given in Table 3.13.
The second a n d less a c c u r a t e m e t h o d to convert TBP from 10
to 760 m m Hg is t h r o u g h Eq. (3.32), w h i c h in its reverse form
b e c o m e s T (760 m m H g ) = 1.17T (10 m m H g ) + 67.51. Esti-
m a t e d TBP at 760 m m Hg t h r o u g h this relation is p r e s e n t e d
in the last c o l u m n of Table 3.13. #
3.2.3 Prediction of Complete Distillation Curves
In m a n y cases distillation d a t a for the entire r a n g e of p e r c e n t
distilled are not available. This is p a r t i c u l a r l y the case w h e n a
fraction contains heavy c o m p o u n d s t o w a r d the e n d of distilla-
tion curve. F o r such fractions distillation can be p e r f o r m e d to
a certain t e m p e r a t u r e . F o r example, in a TBP o r ASTM curve,
distillation d a t a m a y be available at 10, 30, 50, a n d 70% p o i n t s
h u t not at 90 o r 95% points, w h i c h are i m p o r t a n t for process
engineers a n d are characteristics of a p e t r o l e u m p r o d u c t . F o r
heavier fractions the distillation curves m a y even end at 50%
point. F o r such fractions it is i m p o r t a n t t h a t values of temper-
atures at these high p e r c e n t a g e p o i n t s to be e s t i m a t e d f r o m
available data. In this section a d i s t r i b u t i o n function for b o t h
boiling p o i n t a n d density of p e t r o l e u m fractions is p r e s e n t e d
so that its p a r a m e t e r s can be d e t e r m i n e d f r o m as few as three
d a t a p o i n t s on the curve. The function can p r e d i c t the boiling
p o i n t for the entire range from initial p o i n t to 95% point. This
function was p r o p o s e d b y Riazi [31] b a s e d on a p r o b a b i l i t y
d i s t r i b u t i o n m o d e l for the p r o p e r t i e s of h e p t a n e plus fractions
in crude oils a n d reservoir fluids a n d its detailed characteris-
tics are discussed in Section 4.5.4. The d i s t r i b u t i o n m o d e l is
p r e s e n t e d b y the following e q u a t i o n (see Eq. 4.56):
in w h i c h T is the t e m p e r a t u r e on the distillation curve in
kelvin a n d x is the volume o r weight fraction of the m i x t u r e
distilled. A, B, a n d To are the three p a r a m e t e r s to be deter-
m i n e d from available d a t a on the distillation curve t h r o u g h a
l i n e a r regression. To is in fact the initial boiling p o i n t (T at x =
0) b u t has to be d e t e r m i n e d from actual d a t a with x > 0. The
e x p e r i m e n t a l value of To s h o u l d n o t be i n c l u d e d in the regres-
sion process since it is n o t a reliable point. E q u a t i o n (3.35)
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S 109

Superatmospheric

I
ASTMD2887
Simulated EFV
Distillation (P >760 mmHg)
(SD)

ASTM D86 EFV

760 mmHg 760 mmHg

TBP ! --'-I
TBP
1, 50, 100mm 6 760 mmHg
Iv

TBP EFV
10 mmHg 10 mrnHg

ASTMD-1160
10 mmHg

4 V

ASTMD-1160 ASTMD-1160
6 Reported
at
1, 30, 50 mmHg ~" 760 mmHg

STEP METHOD A METHOD B

Eqs. (3.14)or(3.15) Eqs. (3.20)- (3.22)
Eq. (3.34)
Eqs. (3.18) & (3.19) Eqs. (3.26)- (3.28)
Eq. (3.31) Eq. (3.32)
Eq. (3.16)
Eq. (3.29)
Eqs. (3.23)- (3.25)
FIG. 3.20--Summary of methods for the interconversion of various distillation curves.

does not give a finite value for T at x = 1 (end point at 100%
distilled). According to this model the final boiling point is
infinite (oo), which is true for heavy residues. Theoretically,
even for light products with a limited boiling range there is a
very small a m o u n t of heavy c o m p o u n d since all c o m p o u n d s
in a mixture cannot be completely separated by distillation.
For this reason predicted values from Eq. (3.35) are reliable
up to x = 0.99, but not at the end point. Parameters A, B, and
To in Eq. (3.35) can be directly determined by using Solver
(in Tools) in Excel spreadsheets. Another way to determine
the constants in Eq. (3.35) is t h r o u g h its conversion into the
following linear form:
(3.36) Y = Cz + C2X
where Y = In [(T - To) / To] and X - - In In [1 / (1 - x ) ] . Con-
stants C1 and Ca are determined from linear regression of Y
versus X with an initial guess for To. Constants A and B are
determined from C1 and C2 as B = 1/C2 and A = B exp (C1B).
Parameter To can be determined by several estimates to maxi-
mize the R squared (RS) value for Eq. (3.36) and minimize the
AAD for prediction of T form Eq. (3.35). If the initial boiling
point in a distillation curve is available it can be used as the
110 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

TABLE 3.13---Conversion of ASTM D 1160 to TBP at 760 mm Hg Development of these relations is discussed in C h a p t e r 4.
for the petroleum fraction of Example 3.6. I n Eq. (3.35), i f x is v o l u m e fraction, t h e n Tar c a l c u l a t e d f r o m
XZol% ASTM
distilled D 1160, ~ TBP%,~ TBl~76O,
~ TBP~76o,~ Eq. (3.37) w o u l d be volume average boiling p o i n t (VABP)
a n d if x is the weight fraction t h e n Tar w o u l d b e equivalent
10 150 142.5 280.1 280.7
30 205 200.9 349.9 349.0 to the weight average boiling p o i n t (WABP). S i m i l a r l y m o l e
50 250 250 407.2 406.5 average boiling p o i n t c a n be e s t i m a t e d from this e q u a t i o n
70 290 290 453.1 453.3 if x is in m o l e fraction. However, the m a i n a p p l i c a t i o n of
90 350 350 520.4 523.5 Eq. (3.35) is to p r e d i c t c o m p l e t e distillation curve from a lim-
aEq. (3.34).
bEq. (3.29). ited d a t a available. It can also be used to p r e d i c t boiling p o i n t
eEq. (3.32). of residues in a crude oil as will be s h o w n in C h a p t e r 4. Equa-
tion (3.35) is also perfectly a p p l i c a b l e to density o r specific
initial guess, b u t value of To s h o u l d always be less t h a n value gravity d i s t r i b u t i o n along a distillation curve for a p e t r o l e u m
of T for the first d a t a point. F o r fractions with final boiling fraction a n d c r u d e oils. F o r the case of density, p a r a m e t e r T
p o i n t very high a n d uncertain, such as a t m o s p h e r i c o r vac- is r e p l a c e d b y d or SG and density of the m i x t u r e m a y be cal-
u u m residues a n d h e p t a n e - p l u s fraction of crude oils, value o f c u l a t e d from Eq. (3.37). W h e n Eqs. (3.35)-(3.38) are u s e d for
B c a n be set as 1.5 a n d Eq. (3.35) reduces to a t w o - p a r a m e t e r p r e d i c t i o n of d e n s i t y of p e t r o l e u m fractions, the value of RS
equation. However, for various p e t r o l e u m fractions with fi- is less t h a n t h a t of distillation data. While the value of B for
nite boiling r a n g e p a r a m e t e r B s h o u l d be d e t e r m i n e d from the case of density is greater t h a n that of boiling p o i n t a n d is
the regression analysis a n d value of B for light fractions is usually 3 for very heavy fractions (C7+) a n d h i g h e r for lighter
h i g h e r t h a n t h a t of heavier fractions a n d is n o r m a l l y g r e a t e r mixtures. It s h o u l d be n o t e d t h a t w h e n Eqs. (3.35)-(3.38) are
t h a n 1.5. E q u a t i o n (3.35) can be a p p l i e d to a n y type of distil- a p p l i e d to specific gravity o r density, x s h o u l d b e cumulative
lation data, ASTM D 86, ASTM D 2887 (SD), TBP, EFV, a n d v o l u m e fraction. F u r t h e r p r o p e r t i e s a n d a p p l i c a t i o n of this
ASTM D 1160 as well as TBP at r e d u c e d pressures o r EFV d i s t r i b u t i o n function as well as m e t h o d s of calculation of av-
at elevated pressures. I n the case of SD curve, x is c u m u l a - erage p r o p e r t i e s for the m i x t u r e are given in C h a p t e r 4. Here-
tive weight fraction distilled. The average boiling p o i n t of the in we d e m o n s t r a t e use for this m e t h o d for p r e d i c t i o n of dis-
fraction c a n be d e t e r m i n e d from the following relation: tillation curves of p e t r o l e u m fractions t h r o u g h the following
example.
Tav -= To(1 + Tar)
(3.37' (A):~ ( 1) Example 3 . 7 - - A S T M D 86 distillation d a t a from initial to
T~*v= F 1+
final boiling p o i n t for a gas oil s a m p l e [ i] are given in the first
two c o l u m n s of Table 3.14. Predict the distillation curve for
in w h i c h F is the g a m m a function a n d m a y be d e t e r m i n e d
the following four cases:
f r o m the following relation w h e n value of p a r a m e t e r B is
g r e a t e r t h a n 0.5.
a. Use d a t a points at 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, a n d
F (l +l) =o.992814-O.504242B-l +O.696215B -z 95 vo1% distilled.
b. Use all d a t a p o i n t s from 5 to 70 vol% distilled.
- 0 . 2 7 2 9 3 6 B -3 + 0.088362B -4 c. Use three d a t a p o i n t s at 10, 30, a n d 50%.
(3.38) d. Use three d a t a p o i n t s at 30, 50, a n d 70%.

TABLE 3.14--Prediction of ASTM D 86 distillation curve for gas oil sample of Example 3.7.
Data Set A Data Set B Data Set C Data Set D
Vol% distilled Temp.exp, K Pred, K AD, K Pred, K AD, K Pred, K AD, K Pred, K AD, K
0 520.4 526.0 5.6 525.0 4.6 530.0 9.6 512.0 8.4
5 531.5 531.6 0.1 531.5 0.0 532.7 1.2 526.4 5.1
10 534.8 534.6 0.2 534.7 0.1 534.8 0.0 531.2 3.6
20 539.8 539.5 0.4 539.6 0.2 538.9 0.9 537.8 2.0
30 543.2 543.8 0.7 544.0 0.8 543.1 0.0 543.1 0.0
40 548.2 548.1 0.1 548.1 0.0 547.6 0.5 547.9 0.3
50 552.6 552.5 0.1 552.4 0.2 552.6 0.0 552.6 0.0
60 557.0 557.3 0.3 557.0 0.1 558.5 1.4 557.4 0.4
70 562.6 562.9 0.3 562.2 0.4 565.6 3.0 562.6 0.0
80 570.4 569.9 0.5 568.7 1.7 575.2 4.8 568.8 1.6
90 580.4 580.4 0.0 578.2 2.2 590.6 10.3 577.5 2.9
95 589.8 589.6 0.2 586.6 3.3 605.2 15.4 584.8 5.1
100 600.4 608.3 7.9 603.1 2.7 637.1 36.7 598.4 2.0
AAD (total), K 1.3 1.3 6.5 2.4
No. of data used 11 8 3 3
To 526 525 530 512
A 0.01634 0.0125 0.03771 0.00627
B 1.67171 1.80881 1.21825 2.50825
RS 0.9994 0.999 1 1
AAD (data used), K 0.25 0.23 0 0
VABP, K 554.7 555.5 555 557 554

For each case give parameters To, A, and B in Eq. (3.35) as
well as value of RS and AAD based on all data points and
based on data used for the regression. Also calculate VABP
from Eq. (3.37) and compare with actual VABP calculated
from Eq. (3.6).
Solution--Summary of calculation results for all four cases
are given in Table 3.14. For Case A all experimental data given
on the distillation curve (second column in Table 3.14) from
5 to 95% points are used for the regression analysis by Eq.
(3.36). Volume percentages given in the first column should be
converted to cumulative volume fraction, x, (percent values
divided by 100) and data are converted to X and Y defined in
Eq. (3.36). The first data point used in the regression process
is at x = 0.05 with T = 531.5 K; therefore, the initial guess
(To) should be less than 531. With a few changes in To values,
the maximum RS value of 0.9994 is obtained with minimum
AAD of 0.25 K (for the 11 data points used in the regression
process). The AAD for the entire data set, including the IBP
and FBP, is 1.3 K. As mentioned earlier the experimentally re-
ported IBP and especially the value of FBP are not accurate.
Therefore, larger errors for prediction of IBP and FBP are ex-
pected from Eq. (3.35). Since values of FBP at x = 1 are not
finite, the value of T at x = 0.99 may be used as an approx-
imate predicted value of FBP from the model. These values
are given in Table 3.14 as predicted values for each case at
100 vol% vaporized. Estimated VABP from Eq. (3.37) for Case
A is 555.5 versus value of 554.7 from actual experimental data
and definition of VABP by Eq. (3.6).
For Case B, data from 5 to 70 vol% distilled are used for
the regression process and as a result the predicted values
up to 70% are more accurate than values above 70% point.
However, the overall error (total AAD) is the same as for Case
A at 1.3 K. For Case C only three data points at 10, 30, and
50% are used and as a result much larger errors especially
for points above 50% are observed. In Case D, data at 30, 50,
650
o Exp. data
-- Pred.(data set A)
. . . . . . . Pred. (data set B)
600
..... Pred. (data set C)
m ~ Perd. (data set D)
~9
550
500 . . . . . . . .
0 20 40 60
Vol% Distilled
FIG. 3.211Prediction of distillation curves for the gas oil
sample of Example 3.7.
3. CHARACTERIZATION OF PETROLEUM FRACTIONS 111
and 70% points are used and the predicted values are more
accurate than values obtained in Case C. However, for this last
case the highest error for the IBP is obtained because the first
data point used to obtain the constants is at 30%, which is far
from 0% point. Summary of results for predicted distillation
curves versus experimental data are also shown in Fig. 3.2 I.
As can be seen from the results presented in both Table 3.14
and Fig. 3.21, a good prediction of the entire distillation curve
is possible through use of only three data points at 30, 50, and
70%. r
3.3 P R E D I C T I O N OF P R O P E R T I E S
OF P E T R O L E U M F R A C T I O N S
As discussed in Chapter 1, petroleum fractions are mixtures of
many hydrocarbon compounds from different families. The
most accurate method to determine a property of a mixture
is through experimental measurement of that property. How-
ever, as this is not possible for every petroleum mixture, meth-
ods of estimation of various properties are needed by process
or operation engineers. The most accurate method of esti-
mating a property of a mixture is through knowledge of the
exact composition of all components existing in the mixture.
Then properties of pure components such as those given in
Tables 2.1 and 2.2 can be used together with the composi-
tion of the mixture and appropriate mixing rules to determine
properties of the mixture. If experimental data on properties
of pure compounds are not available, such properties should
be estimated through the methods presented in Chapter 2.
Application of this approach to defined mixtures with very
few constituents is practical; however, for petroleum mixtures
with many constituents this approach is not feasible as the
determination of the exact composition of all components
in the mixture is not possible. For this reason appropriate
models should be used to represent petroleum mixtures by
some limited number of compounds that can best represent
the mixture. These limited compounds are different from the
real compounds in the mixture and each is called a "pseudo-
component" or a "pseudocompound". Determination of these
pseudocompounds and use of an appropriate model to de-
"l scribe a mixture by a certain number of pseudocompounds
'1/ is an engineering art in prediction of properties of petroleum
; "t mixtures and are discussed in this section.
1'~'
~o4'(.~q? 3.3.1 M a t r i x o f P s e u d o c o m p o n e n t s Table
As discussed in Chapter 2, properties of hydrocarbons vary
by both carbon number and molecular type. Hydrocarbon
properties for compounds of the same carbon number vary
from paraffins to naphthenes and aromatics. Very few frac-
tions may contain olefins as well. Even within paraffins fam-
ily properties of n-paraffins differ from those of isoparaffins.
Boiling points of hydrocarbons vary strongly with carbon
number as was shown in Table 2.1; therefore, identification of
hydrocarbons by carbon number is useful in property predic-
80 100 tions. As discussed in Section 3,1.5.2, a combination of GS-
MS in series best separate hydrocarbons by carbon number
and molecular type. If a mixture is separated by a distillation
column or simulated distillation, each hydrocarbon cut with a
single carbon number contains hydrocarbons from different
112 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
TABLE 3.15--Presentation of a petroleum fraction (dieselfuel) by a matrix of 30 pseudocomponents.
Carbon number n-Paraffins Isoparaffins
C11 1 2
C12 6 7
C13 11 12
C14 16 17
C15 21 22
C16 26 27
groups, which can be identified by a PIONA analyzer. As an
example in Table 1.3 (Chapter 1), carbon number ranges for
different petroleum products are specified. For a diesel fuel
sample, carbon number varies from Cll to C16 with a boiling
range of 400-550~ If each single carbon number hydrocar-
bon cut is further separated into five pseudocomponents from
different groups, the whole mixture may be represented by a
group of 30 pseudocomponents as shown in Table 3.15. Al-
though each pseudocomponent is not a pure hydrocarbon but
their properties are very close to pure compounds from the
same family with the same carbon number. If the amounts of
all these 30 components are known then properties of the mix-
ture may be estimated quite accurately. This requires exten-
sive analysis of the mixture and a large computation time for
estimation of various properties. The number of pseudocom-
ponents may even increase further if the fraction has wider
boiling point range such as heptane plus fractions as will be
discussed in Chapter 4. However, many petroleum fractions
are olefin free and groups of n-paraffins and isoparaffins may
be combined into a single group of paraffins. Therefore, the
number of different families reduces to three (paraffins, naph-
thenes, and aromatics). In this case the number of compo-
nents in Table 3.15 reduces to 6 x 3 or 18. If a fraction is nar-
row in boiling range then the number of rows in Table 3.15
decreases indicating lower carbon number range. In Chapter
4, boiling points of various single carbon number groups are
given and through a TBP curve it would be possible to deter-
mine the range of carbon number in a petroleum fraction. In
Table 3.15, if every two carbon number groups and all paraf-
fins are combined together, then the whole mixture may be
represented by 3 x 3 or 9 components for an olefin-free frac-
tion. Similarly if all carbon numbers are grouped into a single
carbon number group, the mixture can be represented by only
three pseudocomponents from paraffins (P), naphthenes (N),
and aromatics (A) groups all having the same carbon number.
This approach is called pseudocomponent technique.
Finally the ultimate simplicity is to ignore the difference in
properties of various hydrocarbon types and to present the
whole mixture by just a single pseudocomponent, which is
the mixture itself. The simplicity in this case is that there is
no need for the composition of the mixture. Obviously the
accuracy of estimated properties decreases as the number of
pseudocomponents decreases. However, for narrow boiling
range fractions such as a light naphtha approximating the
mixture with a single pseudocomponent is more realistic and
more accurate than a wide boiling range fraction such as an
atmospheric residuum or the C7+ fraction in a crude oil sam-
ple. As discussed in Chapter 2, the differences between prop-
erties of various hydrocarbon families increase with boiling
point (or carbon number). Therefore, assumption of a single
pseudocomponent for a heavy fraction (M > 300) is less ac-
curate than for the case of light fractions. For fractions that
Olefins Naphthenes Aromatics
3 4 5
8 9 10
13 14 15
18 19 20
23 24 25
28 29 30
are rich in one hydrocarbon type such as coal liquids that
may have up to 90% aromatics, it would be appropriate to di-
vide the aromatics into further subgroups of monoaromatics
(MA) and polyaromatics (PA). Therefore, creation of a matrix
of pseudocomponents, such as Table 3.15, largely depends on
the nature and characteristics of the petroleum mixture as
well as availability of experimental data.
3.3.2 Narrow Versus Wide Boiling Range Fractions
In general, regardless of molecular type, petroleum fractions
may be divided into two major categories: narrow and wide
boiling range fractions. A narrow boiling range fraction was
defined in Section 3.2.1 as a fraction whose ASTM 10-90%
distillation curve slope (SL) is less than 0.8~ although
this definition is arbitrary and may vary from one source to
another. Fractions with higher 10-90% slopes may be consid-
ered as wide boiling range. However, for simplicity the meth-
ods presented in this section for narrow fractions may also
be applied to wider fractions. For narrow fractions, only one
carbon number is considered and the whole fraction may be
characterized by a single value of boiling point or molecu-
lar weight. For such fractions, if molecular type is known
(PNA composition), then the number of pseudocomponents
in Table 3.15 reduces to three and if the composition is not
known the whole mixture may be considered as a single pseu-
docomponent. For this single pseudocomponent, properties
of a pure component whose characteristics, such as boiling
point and specific gravity, are the same as that of the fraction
can be considered as the mixture properties. For mixtures the
best characterizing boiling point is the mean average boiling
point (MeABP); however, as mentioned in Section 3.2.1, for
narrow fractions the boiling point at 50 vol% distilled may
be considered as the characteristic boiling point instead of
MeABP. The specific gravity of a fraction is considered as the
second characteristic parameter for a fraction represented by
a single pseudocomponent. Therefore, the whole mixture may
be characterized by its boiling point (Tb) and specific gravity
(SG). In lieu of these properties other characterization pa-
rameters discussed in Chapter 2 may be used.
Treatment of wide boiling range fractions is more compli-
cated than narrow fractions as a single value for the boiling
point, or molecular weight, or carbon number cannot rep-
resent the whole mixture. For these fractions the number of
constituents in the vertical columns of Table 3.15 cannot be
reduced to one, although it is still possible to combine var-
ious molecular types for each carbon number. This means
that the minimum number of constituents in Table 3.14 for
a wide fraction is six rather than one that was considered
for narrow fractions. The best example of a wide boiling
range fraction is C7+ fraction in a crude oil or a reservoir
fluid. Characterization of such fractions through the use of a
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS 113

distribution model that reduces the mixture into a n u m b e r of The values of To and SGo determined from regression of
p s e u d o c o m p o n e n t s with k n o w n characterization parameters data through Eq. (3.35) do not m a t c h well with the exper-
will be discussed in detail in Chapter 4. However, a simpler imental initial values. This is due to the maximizing value
a p p r o a c h based on the use of TBP curve is outlined in Ref. of RS with data used in the regression analysis. Actually
[32]. In this approach the mixture property is calculated from one can imagine that the actual initial values are lower than
the following relation: experimentally measured values due to the difficulty in such
measurements. However, these initial values do not affect
1
subsequent calculations. Predicted values at all other points
(3.39) 0 = [ O(x) dx from 5 up to 95% are consistent with the experimental values.
1/
0 F r o m calculated values of SGo, A, and B for the SG curve,
in which 0 is the physical property of mixture and O(x) is the one can determine the mixture SG for the whole fraction
value of property at point x on the distillation curve. This t h r o u g h use of Eqs. (3.37) and (3.38). For SG, B = 7.1957 and
approach may be applied to any physical property. The in- from Eq. (3.38), F (I + 1 / B) = 0.9355. From Eq. (3.37) we
get SG~v= ( ~ ) 1 / 7 1 9 5 7 F ( 1 + 7.1~57) ---- 0.5269 x 0.9355 =
tegration should be carried out by a numerical method. The
fraction is first divided into a n u m b e r of p s e u d o c o m p o n e n t s 0.493. Therefore, for the mixture: SGav -- 0.5(1 + 0 . 4 9 3 ) =
along the entire range of distillation curve with k n o w n boiling 0.746. Comparing with experimental value of 0.74, the
points and specific gravity. Then for each c o m p o n e n t physi- percent relative deviation (%D) with experimental value is
cal properties are calculated from methods of Chapter 2 and 0.8%. In Chapter 4 another m e t h o d based on a distribution
finally the mixture properties are calculated through a sim- function is introduced that gives slightly better prediction
ple mixing rule. The procedure is outlined in the following for the density of wide boiling range fractions and crude
example. oils.
To calculate a mixture property such as molecular weight,
the mixture is divided to some narrow pseudocomponents,
Example 3 . 8 - - F o r a low boiling naphtha, TBP curve is pro-
Np. If the mixture is not very wide such as in this exam-
vided along with the density at 20~ as tabulated below [32].
ple, even Nv = 5 is sufficient, but for wider fractions the mix-
Estimate specific gravity and molecular weight of this fraction
ture m a y be divided to even larger n u m b e r of pseudocompo-
using the wide boiling range approach. Compare the calcu-
nents (10, 20, etc.). If Nv = 5, then values of T and SG at x =
lated results with the experimental values reported by Lenior
0, 0.2, 0.4, 0.6, 0.8, and 0.99 are evaluated t h r o u g h Eq. (3.35)
and Hipkin and others [1, 11, 32] as SG = 0.74 and M = 120.
and parameters determined above. Value of x = 0.99 is used
instead o f x = 1 for the end point as Eq. (3.35) is not defined
vo1% 0 (IBP) 5 10 20 30 50 70 90 95
at x = 1. At every point, molecular weight, M, is determined
TBP, K 283.2 324.8 348.7 369.3 380.9 410.4 436.5 467.6 478.7
from methods of Chapter 2. In this example, Eq. (2.50) is quite
d20, g/cm 3 ... 0.654 0.689 0.719 0.739 0.765 0.775 0.775 0.785
accurate and m a y be used to calculate M since all compo-
nents in the mixture have M < 300 (~Nc < 20) and are within
Solution--For this fraction the 10-90% slope based on TPB the range of application of this method. Equation (2.50) is
curve is about 1.49~ This value is slightly above the slope M = 1.6604 x 10-4Tb2"1962 SG -1-~ Calculations are s u m m a -
based on the ASTM D 86 curve but still indicates h o w wide rized in the following table.
the fraction is. For this sample based on the ASTM distilla-
tion data [1], the 10-90% slope is 1.35~ which is above
x Tb, K SG M
the value of 0.8 specified for narrow fractions. To use the
m e t h o d by Riazi-Daubert [32] for this relatively wide frac- 0 240.0 0.500 56.7
0.2 367.1 0.718 99.8
tion, first distribution functions for both boiling point and 0.4 396.4 0.744 114.0
specific gravity should be determined. We use Eqs. (3.35)- 0.6 421.0 0.764 126.7
(3.38) to determine the distribution functions for both prop- 0.8 448.4 0.785 141.6
erties. The molecular weight, M, is estimated for all points on 0.99 511.2 0.828 178.7
the curve t h r o u g h appropriate relations in Chapter 2 devel-
oped for pure hydrocarbons. The value of M for the mixture The trapezoidal rule for integration is quite accurate to
then m a y be estimated from a simple integration over the en- estimate the molecular weight of the mixture. May -- (1/5) x
tire range o f x as given by Eq. (3.39): May = fd M(x)dx, where [(56.7+ 178.7)/2+(99.8 + 114 + 126.7+ 141.6+ 178.7)] =
M(x) is the value of M at point x determined from Tb(X) and 119.96 - 120. This is exactly the same as the experimental
SG(x). May is the average molecular weight of the mixture. value of molecular weight for this fraction [ 1, 11].
For this fraction values of densities given along the distilla- If the whole mixture is considered as a single pseudocom-
tion curve are at 20~ and should be converted to specific ponent, Eq. (2.50) should be applied directly to the mixture
gravity at 15.5~ (60~ through use of Eq. (2.112) in Chap- using the MeABP and SG of the mixture. For this fraction
ter 2: SG = 0.9915d20 + 0.01044. Parameters of Eq. (3.35) for the Watson K is given as 12.1 [1]. F r o m Eq. (3.13) using
both temperature and specific gravity have been determined experimental value of SG, average boiling point is calcu-
and are given as following. lated as Tb ----(12.1 x 0.74)3/1.8 ----398.8 K. F r o m Eq. (2.50),
the mixture molecular weight is 116.1, which is equivalent to
Parameters in Eq. (3.35) To, K SGo A B RS %D = -3.25%. For this sample, the difference between 1 and
TBP curve 240 1.41285 3.9927 0.996 5 p s e u d o c o m p o n e n t s is not significant, but for wider fraction
SG curve 0.5 0.07161 7.1957 0.911 the improvement of the proposed method is m u c h larger.
114 CHARACTERIZATION AND P R O P E R T I E S OF P E T R O L E U M FRACTIONS
Example 3.9--It is assumed for the same fraction of Ex-
ample 3.8, the only information available is ASTM D 86 data:
temperatures of 350.9, 380.9, 399.8, 428.2, and 457.6 K at 10,
30, 50, 70 and 90 vol% distilled, respectively. State how can
you apply the proposed approach for wide boiling range frac-
tions to calculate the molecular weight of the fraction.
Solution--Since distillation data are in terms of ASTM
D 86, the first step is to convert ASTM to TBP through
Eq. (3.14). The second step is to determine the TBP distri-
bution function through Eqs. (3.35) and (3.36). The third
step is to generate values of T at x -- 0, 0.2, 0.4, 0.6, 0.8, and
0.99 from Eq. (3.35) and parameters determined from TBP
distillation curve. Since the specific gravity for this frac-
tion is not known it may be estimated from Eq. (3.17) and
constants in Table 3.4 for the ASTM D 86 data as follows:
SG = 0.08342 • (350.9)~176176 = 0.756. The Wat-
son K is calculated as Kw = (1.8 x 399.8)1/3/0.756 = 11.85.
Now we assume that Kw is constant for the entire range of
distillation curve and on this basis distribution of SG can be
calculated through distribution of true boiling point. At every
point that T is determined from Eq. (3.35) the specific grav-
ity can be calculated as SG : (1.8 x T)I/a/Kw, where T is the
temperature on the TBP curve. Once TBP temperatures and
SG are determined at x = 0, 0.2, 0.4, 0.6, 0.8, and 0.99 points,
molecular weight may be estimated from Eq. (2.50). Numer-
ical integration of Eq. (3.39) can be carried out similar to the
calculations made in Example 3.8 to estimate the molecular
weight. In this approach the result may be less accurate than
the result in Example 3.8, as ASTM distillation curve is used
as the only available data.
Although the method outlined in this section improves
the accuracy of prediction of properties of wide boiling
range fraction, generally for simplicity in calculations most
petroleum products are characterized by a single value of boil-
ing point, molecular weight, or carbon number regardless of
their boiling range. The proposed method is mainly applied
to crude oils and C7+ fraction of reservoir fluids with an ap-
propriate splitting technique as is shown in the next chapter.
However, as shown in the above example, for very wide boiling
range petroleum products the method presented in this sec-
tion may significantly improve the accuracy of the estimated
physical properties.
3.3.3 Use of Bulk Parameters
(Undefined Mixtures)
An undefined petroleum fraction is a fraction whose com-
position (i.e., PNA) is not known. For such fractions infor-
mation on distillation data (boiling point), specific gravity,
or other bulk properties such as viscosity, refractive index,
CH ratio, or molecular weight are needed. If the fraction is
considered narrow boiling range then it is assumed as a sin-
gle component and correlations suggested in Chapter 2 for
pure hydrocarbons may be applied directly to such fractions.
All limitations for the methods suggested in Chapter 2 should
be considered when they are used for petroleum fractions.
As mentioned in Chapter 2, the correlations in terms of Tb
and SG are the most accurate methods for the estimation of
various properties (molecular weight, critical constants, etc.).
For narrow boiling range fractions, ASTM D 86 temperature
at 50 vol% vaporized may be used as the characterizing boil-
ing point for the whole mixture. However, for a wide boiling
range fraction if it is treated as a single pesudocomponent
the MeABP should be calculated and used as the character-
izing parameter for Tb in the correlations of Chapter 2. If
for a fraction TBP distillation data are available the average
boiling point calculated through Eq. (3.37) with parameters
determined from TBP curve would be more appropriate than
MeABP determined from ASTM D 86 curve for use as the
characterizing boiling point. For cases where only two points
on the distillation curve are known the interpolated value at
50% point may be used as the characterizing boiling point of
the fraction.
For heavy fractions (M > 300) in which atmospheric dis-
tillation data (ASTM D 86, SD, or TBP) are not available, if
ASTM D 1160 distillation curve is available, it should be con-
verted to ASTM D 86 or TBP through methods outlined in
Section 3.2. In lieu of any distillation data, molecular weight
or viscosity may be used together with specific gravity to esti-
mate basic parameters from correlations proposed in Chap-
ter 2. If specific gravity is not available, refractive index or
carbon-to-hydrogen weight ratio (CH) may be used as the
second characterization parameter.
3.3.4 Method of Pseudocomponent
(Defined Mixtures)
A defined mixture is a mixture whose composition is known.
For a petroleum fraction if at least the PNA composition
# is known it is called a defined fraction. Huang [11,33] used
the pseudocompounds approach to estimate enthalpies of
narrow and defined petroleum fractions. This technique has
been also used to calculate other physical properties by
other researchers [34, 35]. According to this method all com-
pounds within each family are grouped together as a sin-
gle pseudocomponent. An olefin-free fraction is modeled
into three pseudocomponents from three homologous groups
of n-alkanes (representing paraffins), n-alkylcyclopentanes
or n-alkylcyclohexanes (representing naphthenes), and n-
alkylbenzenes (representing aromatics) having the same boil-
ing point as that of ASTM D 86 temperature at 50% point.
Physical properties of a mixture can be calculated from prop-
erties of the model components by the following mixing rule:
(3.40) 0 : XpOp@XNON+ XAOA
where 0 is a physical property for the mixture and 0p, ON, and
0A are the values of 0 for the model pseudocomponents from
the three groups. In this equation the composition presented
by xA, XN, and XAshould be in mole fraction, but because the
molecular weights of different hydrocarbon groups having
the same boiling point are close to each other, the compo-
sition in weight or even volume fractions may also be used
with minor difference in the results. If the fraction contains
olefinic compounds a fourth term for contribution of this
group should be added to Eq. (3.40). Accuracy of Eq. (3.40)
can be increased if composition of paraffinic group is known
in terms of n-paraffins and isoparaffins. Then another pseudo-
component contributing the isoparaffinic hydrocarbons may
be added to the equation. Similarly, the aromatic part may be
split into monoaromatics and polyaromatics provided their
 3. CHARACTERIZATION OF P E T R O L E U M FRACTIONS
amount in the fraction is known. However, based on our expe-
rience the PNA three-pseudocomponent model is sufficiently
accurate for olefin-free petroleum fractions. For coal liquids
with a high percentage of aromatic content, splitting aromat-
ics into two subgroups may greatly increase the accuracy of
model predictions. In using this method the minimum data
needed are at least one characterizing parameter (Tb or M)
and the PNA composition.
Properties of pseudocomponents may be obtained from in-
terpolation of values in Tables 2.1 and 2.2 to match boiling
point to that of the mixture. As shown in Section 2.3.3, prop-
erties of homologous groups can be well correlated to only
one characterization parameter such as boiling point, molec-
ular weight, or carbon number, depending on the availability
of the parameter for the mixture. Since various properties of
pure homologous hydrocarbon groups are given in terms of
molecular weight by Eq. (2.42) with constants in Table 2.6,
if molecular weight of a fraction is known it can be used di-
rectly as the characterizing parameter. But if the boiling point
is used as the characterizing parameter, molecular weights of
the three model components may be estimated through rear-
rangement of Eq. (2.42) in terms of boiling point as following:
(3.41) Mp = [ 1_____2___[6.98291 _ ln(1070 - Tb)]/3/2/
| 0.02013
(3.42) MN = { ~ [ 6 . 9 5 6 4 9 -- ln(1028 - Tb)]} 3/2
(3.43) MA = / 1 - - - - ~ [ 6 . 9 1 0 6 2 -- ln(1015 -- Tb)]} 3/2
| 0.02247
where Mp, MN, and MA are molecular weights of paraffinic,
naphthenic, and aromatic groups, respectively. Tb is the char-
acteristic boiling point of the fraction. Predicted values of Mp,
MN, and MAversus Tb were presented in Fig. 2.15 in Chapter 2.
As shown in this figure the difference between these molecu-
lar weights increase as boiling point increases. Therefore, the
pseudocomponent approach is more effective for heavy frac-
tions. If ASTM D 86 distillation curve is known the tempera-
ture at 50% point should be used for Tb, but if TBP distillation
data are available an average TBP would be more suitable to
be used for Tb. Once Mr, MN, and MA are determined, they
should be used in Eq. (2.42) to determine properties from cor-
responding group to calculate other properties. The method
is demonstrated in the following example.
Example 3.10---A petroleum fraction has ASTM D 86 50%
temperature of 327.6 K, specific gravity of 0.658, molecular
weight of 78, and PNA composition of 82, 15.5, and 2.5 in
vol% [36]. Estimate molecular weight of this fraction using
bulk properties of Tb and SG and compare with the value esti-
mated from the pseudocomponent method. Also estimate the
mixture specific gravity of the mixture through the pseudo-
component technique and compare the result with the exper-
imental value.
Solution--For this fraction the characterizing parameters
are Tb ----327.6 K and SG = 0.672. To estimate M from these
bulk properties, Eq. (2.50) can be applied since the boiling
point of the fraction is within the range of 40-360~ (~C5-
C22). The results of calculation is M = 85.0, with relative
115
deviation of 9%. From the pseudocomponent approach, Mp,
MN, and MA are calculated from Eqs. (3.41)-(3.43) as 79.8,
76.9, and 68.9, respectively. The mixture molecular weight is
calculated through Eq. (3.40) as M = 0.82 x 79.8 + 0.155 x
76.9 + 0.025 • 68.9 = 79, with relative deviation of 1.3%. If
values of Alp, MN, and MA are substituted in Eq. (2.42) for
the specific gravity, we get SGp = 0.651, SGN = 0.749, and
SGA -----0.895. From Eq. (3.40) the mixture specific gravity is
SG = 0.673, with AD of 2.3%. It should be noted that when
Eq. (3.40) is applied to molecular weight, it would be more
appropriate to use composition in terms of mole fraction
rather than volume fraction. The composition can be con-
verted to weight fraction through specific gravity of the three
components and then to mole fraction through molecular
weight of the components by equations given in Section
1.7.15. The mole fractions are Xmp = 0.785, XmN = 0.177, and
Xmn=0.038 and Eq. (3.39) yields M = 78.8 for the mixture
with deviation of 1%. The difference between the use of
volume fraction and mole fraction in Eq. (3.40) is minor and
within the range of experimental uncertainty. Therefore, use
of any form of composition in terms of volume, weight, or
mole fraction in the pseudocomponent method is reasonable
without significant effect in the results. For this reason, in
most cases the PNA composition of petroleum fractions are
simply expressed as fraction or percentage and they may
considered as weight, mole, or volume.
In the above example the method of pseudocomponent pre-
dicts molecular weight of the fraction with much better ac-
curacy than the use of Eq. (2.50) with bulk properties (%AD
of 1.3% versus 9%). This is the case for fractions that are
highly rich in one of the hydrocarbon types. For this frac-
tion paraffinic content is nearly 80%, but for petroleum frac-
tions with normal distribution of paraffins, naphthenes, and
aromatics both methods give nearly similar results and the
advantage of use of three pseudocomponents from different
groups over the use of single pseudocomponent with mixture
bulk properties is minimal. For example, for a petroleum frac-
tion the available experimental data are [36] M = 170, Tb =
487 K, SG = 0.802, xp --- 0.42, XN = 0.41, andxn = 0.17. Equa-
tion (2.50) gives M = 166, while Eq. (3.40) gives M = 163 and
SG = 0.792. Equation (3.40) is particularly useful when only
one bulk property (i.e., Tb) with the composition of a frac-
tion is available. For highly aromatic (coal liquids) or highly
paraffinic mixtures the method of pseudocomponent is rec-
ommended over the use of bulk properties.
3.3.5 Estimation of Molecular Weight, Critical
Properties, and Acentric Factor
Most physical properties of petroleum fluids are calculated
through corresponding state correlations that require pseu-
docritical properties (Tpc, Pr~, and Vpc) and acentric factor
as a third parameter. In addition molecular weight (M) is
needed to convert calculated mole-based property to mass-
based property. As mentioned in Section 1.3, the accuracy
of these properties significantly affects the accuracy of es-
timated properties. Generally for petroleum fractions these
basic characterization parameters are calculated through ei-
ther the use of bulk properties and correlations of Chapter 2
116 CHARACTERIZATION AND P R O P E R T I E S OF P E T R O L E U M FRACTIONS

TABLE 3.16---Comparison of various methods of predicting pseudocritical properties and

acentric factor through enthalpy calculation of eight petroleum fractions [37].
Method of estimatinginput AAD,kJ/kg
Parameters*~ Liquid Vapor
Item To, Pc to (437 data points) (273 data points)
1 Pseudocomp. Pseudocomp. 5.3 7.9
2 RD (80) LK 5.9 7.7
3 KL LK 5.8 7.4
4 Winn LK 9.9 12.8
~*Pseudocomp.: The pseudocomponent method by Eqs. (3.40)-(3.43) and (2.42) for T~, P~ and co; RD: Riazi-
Daubert [38] by Eqs. (2.63) and (2.64); LK: Lee,-Kesler [39] by Eq. (2.103); KL: Kester-Lee [40] by Eqs. (2.69)
and (2.70); Winn method [41] by Eqs. (2.94) and (2.95),

or by the pseudocomponent approach as discussed in Sec-
tions 3.3.2-3.3.4.
For petroleum fractions, pseudocritical properties are not
directly measurable and therefore it is not possible to make
a direct evaluation of different methods with experimental
data. However, these methods can be evaluated indirectly
through prediction of other measurable properties (i.e., en-
thalpy) through corresponding state correlations. These cor-
relations are discussed in Chapters 6-8. Based on more than
700 data points for enthalpies of eight petroleum fractions
over a wide range of temperature and pressure [1], different
methods of estimation of pseudocritical temperature, pres-
sure (Tpo Pp~), and acentric factor (w) have been evaluated
and compared [37]. These petroleum fractions ranging from
naphtha to gas oil all have molecular weights of less than 250.
Details of these enthalpy calculations are given in Chapter 7.
Summary of evaluation of different methods is given in Ta-
ble 3.16. As shown in Table 3.16, the methods of pseudocom-
ponent, Lee-Kesler, and Riazi-Daubert have nearly similar
accuracy for estimating the critical properties of these light
petroleum fractions. However, for heavier fractions as it is
shown in Example 3.11, the methods of pseudocomponent
provide more accurate results.
Example 3.1 l k E x p e r i m e n t a l data on molecular weight and
composition of five heavy petroleum fractions are given in
Table 3.17, In addition, normal boiling point, specific grav-
ity, density, and refractive index at 20~ are also given [36].
Calculate the molecular weight of these fractions from the fol-
lowing five methods: (1) API method [2, 42] using Eq. (2.51),
(2) Twu method [42] using Eqs. (2.89)-(2.92), (3) Goossens
method [43] using Eq. (2.55), (4) Lee-Kesler method [40] us-
ing Eq. (2.54), and (5) the pseudocomponent method using
Eqs. (3.40)-(3.43). Calculate the %AAD for each method.
Solution--Methods 1, 2, and 4 require bulk properties of Tb
and SG, while the method of pseudocomponent requires Tb
and the PNA composition as it is shown in Example 3.10.
Method 2 requires Tb and density at 20~ (d20). Results of
calculations are given in Table 3.18.
The Twu method gives the highest error (AAD of 14.3%) fol-
lowed by the Goossens with average deviation of 11.4%. The
Twu and Goossens methods both underestimate the molecu-
lar weight of these heavy fractions. The Lee-Kesler method is
more accurate for lighter fractions, while the API method is
more accurate for heavier fractions. The pseudocomponent
method gives generally a consistent error for all fractions and
the lowest AAD%. Errors generated by the API, Lee-Kesler,
and the pseudocomponent methods are within the experi-
mental uncertainty in the measurement of molecular weight
of petroleum fractions. r
In summary, for light fractions (M < 300) methods recom-
mended by the API for Tr and PC (Eqs. 2.65 and 2.66) [2] or the
simple method of Riazi-Daubert (Eqs. 2.63 and 2.64) [38] are
suitable, while for heavier fractions the Lee-Kesler method
(Eqs. 2.69 and 2.70) [40] may be used. The pseudocompo-
nent method may also be used for both Tc and Pc when the
composition is available. For all fractions methods of calcu-
lation of acentric factor from the pseudocomponent or the
method of Lee-Kesler [39] presented by Eq. (2.105) may be
used. Molecular weight can be estimated from the API method
[2] by Eq. (2.51) from the bulk properties; however, if the PNA
composition is available the method of pseudocomponent is
preferable especially for heavier fractions.
3.3.6 E s t i m a t i o n o f D e n s i t y , S p e c i f i c G r a v i t y ,
Refractive Index, and Kinematic Viscosity
Density (d), specific gravity (SG), and refractive index (n) are
all bulk properties directly measurable for a petroleum mix-
ture with relatively high accuracy. Kinematic viscosity at 37.8
or 98.9~ (1)38(100), 1,'99(210))are usually reported for heavy frac-
tions for which distillation data are not available. But, for
light fractions if kinematic viscosity is not available it should
be estimated through measurable properties. Methods of es-
timation of viscosity are discussed in Chapter 8; however, in
this chapter kinematic viscosity at a reference of temperature
of 37.8 or 98.9~ (100~ or 210~ is needed for estimation
of viscosity gravity constant (VGC), a parameter required for
prediction of composition of petroleum fractions. Generally,

TABLE3.17--Molecular weight and composition of five heavy petroleum fractions of Example 3.11 [36].
No. M Tb,~ SG d20, g/ml n20 P% N% A%
1 233 298.7 0.9119 0.9082 1.5016 34.1 45.9 20.0
2 267 344.7 0.9605 0.9568 1.5366 30.9 37.0 32.1
3 325 380.7 0.8883 0.8845 1.4919 58.4 28.9 12.7
4 403 425.7 0.9046 0.9001 1.5002 59.0 28.0 13.0
5 523 502.8 0.8760 0.8750 1.4865 78.4 13.3 8.3
 3. C H A R A C T E R I Z A T I O N O F P E T R O L E U M F R A C T I O N S 117

TABLE 3.18---Comparison of various methods of predicting molecular weight of petroleum fractions of Table 3.17 (Example 3.12).
(1) API (2) Twu, (3) Goossens, (4) Lee-Kesler, (5) Pseudocomp.,
Eq. (2.51) Eqs. (2.89)-(2.92) Eq. (2.55) Eq. (2.54) Eqs. (3.40)-(3.43)
No. M, exp, M, calc AD% M, catc AD% M, calc AD% M, calc AD% M, calc AD%
1 233 223.1 4.2 201.3 13.6 204.6 12.2 231.7 0.5 229.1 1.7
2 267 255.9 4.2 224.0 16.1 235.0 12.0 266.7 0.1 273.2 2.3
3 325 320.6 1.4 253.6 16.8 271.3 11.0 304.3 0.2 321.9 1,0
4 403 377.6 6.3 332.2 17.6 345.8 14.2 374.7 7.0 382.4 5.1
5 523 515.0 1.5 485.1 7.2 483.8 7.5 491,7 6.0 516.4 1.3
Total, AAD% 3.5 14.3 11.4 2.8 2.3

density, which is required in various predictive methods mea-
sured at 20~ is shown by d or d20 in g/mL. These properties
can be directly estimated through bulk properties of mixtures
using the correlations provided in Chapter 2 with good accu-
racy so that there is no need to use the pseudocomponent
approach for their estimation.
Specific gravity (SG) of petroleum fractions may be es-
timated from methods presented in Sections 2.4.3 and 2.6.
If API gravity is known, the specific gravity should be di-
rectly calculated from definition of API gravity using Eq. (2.5).
If density at one temperature is available, then Eq. (2.110)
should be used to estimate the increase in density with de-
crease in temperature and therefore density at 15.5~ (60~
may be calculated from the available density. A simpler re-
lation between SG and da0 based on the rule of thumb is
SG = 1.005d20. If density at 20~ (d20) is available, the fol-
lowing relation developed in Section 2.6.1 can be used:
(3.44) SG -- 0.01044 + 0.9915d20
where d20 is in g/mL and SG is the specific gravity at 15.5~
This equation is quite accurate for estimating density of
petroleum fractions. If no density data are available, then
SG may be estimated from normal boiling point and refrac-
tive index or from molecular weight and refractive index for
heavy fractions in which boiling point may not be available.
Equation (2.59) in Section 2.4.3 gives SG from Tb and I for
fractions with molecular weights of less than 300, while for
heavier fractions Eq. (2.60) can be used to estimate specific
gravity from M and I. Parameter I is defined in terms of re-
fractive index at 20~ n20, by Eq. (2.36). If viscosity data are
available Eq. (2.61) should be used to estimate specific grav-
ity, and finally, if only one type of distillation curves such as
ASTM D 86, TBP, or EFV data are available Eq. (3.17) may be
used to obtain the specific gravity.
Density (d) of liquid petroleum fractions at any temper-
ature and atmospheric pressure may be estimated from the
methods discussed in Section 2.6. Details of estimation of
density of petroleum fractions are discussed in Chapter 6;
however, for the characterization methods discussed in this
chapter, at least density at 20~ is needed. If specific gravity
is available then the rule of thumb with d = 0.995SG is the
simplest way of estimating density at 20~ For temperatures
other than 20~ Eq. (2.110) can be used. Equation (2.113)
may also be used to estimate d20 from Tb and SG for light
petroleum fractions (M < 300) provided that estimated den-
sity is less than the value of SG used in the equation. This
is an accurate way of estimating density at 20~ especially
for light fractions. However, the simplest and most accurate
method of estimating d20 from SG for all types of petroleum
fractions is the reverse form of Eq. (3.44), which is equivalent
to Eq. (2.111). If the specific gravity is not available, then it is
necessary to estimate the SG at first step and then to estimate
the density at 20 ~C. The liquid density decreases with increase
in temperature according to the following relationship [24].
d = dlss - k(T - 288.7)
where d15.5 is density at 15.5~ which may be replaced by
0.999SG. T is absolute temperature in kelvin and k is a con-
stant for a specific compound. This equation is accurate
within a narrow range of temperature and it may be applied
to any other reference temperature instead of 15.5~ Value
of k varies with hydrocarbon type; however, for gasolines it is
close to 0.00085 [24].
Refractive index at 20~ n20, is an important character-
ization parameter for petroleum fractions. It is needed for
prediction of the composition as well as estimation of other
properties of petroleum fractions. If it is not available, it may
be determined from correlations presented in Section 2.6 by
calculation of parameter I. Once I is estimated, n20 can be cal-
culated from Eq. (2.114). For petroleum fractions with molec-
ular weights of less than 300, Eq. (2.115) can be used to esti-
mate I from Tb and SG [38]. A more accurate relation, which
is also included in the API-TDB [2], is given by Eq. (2.116).
For heavier fraction in which Tb may not be available, Eq.
(2.117) in terms of M and SG may be used [44]. Most re-
cently Riazi and Roomi [45] made an extensive analysis of
predictive methods and application of refractive index in pre-
diction of other physical properties of hydrocarbon systems.
An evaluation of these methods for some petroleum fractions
is demonstrated in the following example.
E x a m p l e 3.12--Experimental data on M, Tb, SG, d2o, and n20
for five heavy petroleum fractions are given in Table 3.17, Es-
timate SG, dzo, and n2o from available methods and calculate
%AAD for each method with necessary discussion of results.
S o l u t i o n - - T h e first two fractions in Table 3.17 may be consid-
ered light (M < 300) and the last two fractions are considered
as heavy (M > 300). The third fraction can be in either cat-
egory. Data available on d20, Tb, n20, and M may be used to
estimate specific gravity. As discussed in this section, SG can
be calculated from d20 or from Tb and / or from M and I.
In this example specific gravity may be calculated from four
methods: (1) rule of thumb using d20 as the input parameter;
(2) from d20by Eq. (3.44); (3) from Tb and n20 by Eq. (2.59); (4)
from M and n20 by Eq. (2.60). Summary of results is given in
Table 3.19. Methods 1 and 2, which use density as the input
parameter, give the best results. Method 4 is basically devel-
oped for heavy fractions with M > 300 and therefore for the
last three fractions density is predicted with better accuracy.
1 18 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

TABLE 3.19--Comparison of various methods of predicting specific gravity of petroleum fractions of Table 3.17 (Example 3.12).
(1) Rule of thumb, (2) Use of d20, (3) Use of Tb & n20, (4) Use of M & n20,
SG = 1.005 d20 Eq. (3.44) Eq. (2.59) for M < 300 Eq. (2.60) for M >300
No. M, exp SG, exp SG, calc AD% SG, calc AD% SG, calc AD% SG, calc AD%
1 233 0.9119 0.9127 0.09 0.9109 0.11 0.8838 3.09 0.8821 3.27
2 267 0.9605 0.9616 0.11 0.9591 0.15 0.9178 4.44 0.9164 4.59
3 325 0.8883 0.8889 0.07 0.8874 0.10 0.8865 0.20 0.8727 1.76
4 403 0.9046 0.9046 0.00 0.9029 0.19 0.9067 0.23 0.8867 1.98
5 523 0.8760 0.8794 0.39 0.8780 0.23 0.9062 3.45 0.8701 0.67
Total, AAD% 0.13 0.15 2.28 2.45

M e t h o d 3, w h i c h is r e c o m m e n d e d for light fractions, gives K i n e m a t i c v i s c o s i t y of p e t r o l e u m fractions can be esti-

b e t t e r results for the specific gravity of heavy fractions. It
should he n o t e d t h a t the b o u n d a r y of 300 for light a n d heavy
fractions is a p p r o x i m a t e a n d m e t h o d s p r o p o s e d for light frac-
tions can be u s e d well above this b o u n d a r y limit as s h o w n in
M e t h o d 3.
E s t i m a t i o n of density is s i m i l a r to e s t i m a t i o n o f specific
gravity. W h e n b o t h Tb a n d SG are available Eq. (2.113) is the
m o s t a c c u r a t e m e t h o d for e s t i m a t i o n of density of p e t r o l e u m
fractions. This m e t h o d gives AAD of 0.09% for the five frac-
tions of Table 3.17 with h i g h e r errors for the last two fractions.
This equation m a y be u s e d safely up to m o l e c u l a r weight of
500 b u t for heavier fractions Eq. (3.44) o r the rule of t h u m b
s h o u l d be used. Predicted value of density at 20~ from Eq.
(2.113) is n o t reliable if it is greater t h a n the value of specific
gravity used in the equation. The m e t h o d of rule of t h u m b
with d = 0.995 SG gives an AAD of 0.13% a n d Eq. (3.44) gives
an AAD of 0.15%.
Refractive index is e s t i m a t e d from three different m e t h o d s
a n d results are given in Table 3.20. I n the first m e t h o d , Tb
a n d SG are u s e d as the i n p u t p a r a m e t e r s with Eq. (2.115) to
e s t i m a t e I a n d n is calculated from Eq. (2.114). In the sec-
o n d m e t h o d Eq. (2.116) is u s e d with the s a m e i n p u t data.
E q u a t i o n s (2.115) and (2.116) are b o t h developed with d a t a
on refractive index of p u r e h y d r o c a r b o n s w i t h M < 300. How-
ever, Eq. (2.116) in this range of a p p l i c a t i o n is m o r e a c c u r a t e
t h a n Eq. (2.115). But for heavier fractions as s h o w n in Table
3.20, Eq. (3.115) gives better result. This is due to the simple
n a t u r e of Eq. (2.115) w h i c h allows its a p p l i c a t i o n to heavier
fractions. E q u a t i o n (2.116) does not give very a c c u r a t e refrac-
tive index for fraction with m o l e c u l a r weights of 500 o r above.
E q u a t i o n (2.117) in t e r m s of M a n d SG is developed basically
for heavy fractions a n d for this r e a s o n it does n o t give accu-
rate results for fractions with m o l e c u l a r weights of less t h a n
300. This m e t h o d is p a r t i c u l a r l y useful w h e n boiling p o i n t is
not available b u t m o l e c u l a r weight is available or estimable.
However, if boiling p o i n t is available, even for heavy fractions
Eq. (2.115) gives m o r e a c c u r a t e results t h a n does Eq. (2.117)
as s h o w n in Table 3.20.
m a t e d from m e t h o d s p r e s e n t e d in Section 2.7 of the previous
chapter. At reference t e m p e r a t u r e s of 37.8 a n d 98.9~ (100
a n d 210~ v3m00) a n d 1)99(210) can be d e t e r m i n e d f r o m Eqs.
(2.128) a n d (2.129) or t h r o u g h Fig. 2.12 using API gravity a n d
Kw as the i n p u t p a r a m e t e r s . In use of these equations atten-
tion s h o u l d be p a i d to the l i m i t a t i o n s a n d to check if API a n d
Kw are within the ranges specified for the m e t h o d . To calcu-
late k i n e m a t i c viscosity at a n y o t h e r t e m p e r a t u r e , Eq. (2.130)
o r Fig. 2.13 m a y be used. The p r o c e d u r e is best d e m o n s t r a t e d
t h r o u g h the following example.
Example 3.13--A p e t r o l e u m fraction is p r o d u c e d t h r o u g h
distillation of a Venezuelan crude oil a n d has the specific grav-
ity of 0.8309 a n d the following ASTM D 86 distillation data:
vol% distilled 10 30 50 70 90
ASTM D 86 temperature,~ 423 428 433 442 455
E s t i m a t e k i n e m a t i c viscosity of this fraction at 100 a n d 140~
(37.8 a n d 60~ C o m p a r e the calculated values with the ex-
p e r i m e n t a l values of 1.66 a n d 1.23 cSt [46].
Solution--Kinematic viscosities at 100 a n d 210~ v3m00) a n d
v99(210), are calculated from Eqs. (2.128) a n d (2.129), respec-
tively. The API gravity is calculated from Eq. (2.4): API = 38,8.
To calculate Kw from Eq. (2.13), MeABP is required, F o r this
fraction since it is a n a r r o w boiling range the MeABP is n e a r l y
the s a m e as the m i d boiling p o i n t o r ASTM 50% t e m p e r a t u r e .
However, since c o m p l e t e ASTM D 86 curve is available we use
Eqs. (3.6)-(3.12) to e s t i m a t e this average boiling point. Cal-
culated p a r a m e t e r s are VABP = 435.6~ a n d SL = 0.4~
F r o m Eqs. (3.8) a n d (3.12) we get MeABP = 434~ (223.3~
As expected this t e m p e r a t u r e is very close to ASTM 50%
t e m p e r a t u r e of 433~ F r o m Eq. (2.13), Kw = 11.59. Since
0 < A P I < 80 a n d 10 < Kw < 11, we can use Eqs. (2.128)
a n d (2.129) for calculation of k i n e m a t i c viscosity a n d we
get 1238(100) = 1.8, 1299(210) = 0.82 cSt. To calculate viscosity at
r 140 ~ v60(140), we use Eqs. (2,130)-(2.132). F r o m Eq. (2.131)

TABLE 3.20--Comparison of various methods of predicting refractive index of petroleum fractions of Table 3.17
(Example 3.12).
(1) Use of Tb & SG, (2) Use of Tb & SG, (4) Use o f M & SG,
Eq. (2.115) for M < 300 Eq. (2.116) for M < 300 Eq. (2.117) for M > 300
No. M, exp n20 exp n2o exp. AD% n20 calc AD% n2o calc AD%
1 233 1.5016 1.5122 0.70 1.5101 0.57 1.5179 1.08
2 267 1.5366 1.5411 0.29 1.5385 0.13 1.5595 1.49
3 325 1.4919 1.4960 0.28 1.4895 0.16 1.4970 0.34
4 403 1.5002 1.5050 0.32 1.4952 0.34 1.5063 0.41
5 523 1.4865 1.4864 0.01 1.4746 0.80 1.4846 0.13
Total, AAD% 0.32 0.40 0.69
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S
since 1)38(100) > 1.5 a n d 1)99(210) < 1.5 c S t w e have C38(100) -~- 0
and c99(210)= 0.0392. From Eq. (2.132), A1 = 10.4611, B1 =
-4.3573, D1 = -0.4002, DE = --0.7397, and from Eq. (2.130)
at T -- 140~ (60~ we calculate the kinematic viscosity. It
should be noted that in calculation of v60(140)from Eq. (2.130)
trial and error is required for calculation of parameter c. At
first it is assumed that c = 0 and after calculation of 1)60(140)if
it is less than 1.5 cSt, parameter c should be calculated from
Eq. (2.131) and substituted in Eq. (2.130). Results of calcula-
tions are as follows: 1)38(100) = 1.8 and 1)60(140)= 1.27 cSt. Com-
paring with the experimental values, the percent relative de-
viations for kinematic viscosities at 100 and 140~ are 8.4 and
3.3%, respectively. The result is very good, but usually higher
errors are observed for estimation of kinematic viscosity of
petroleum fractions from this method.
3.4 G E N E R A L P R O C E D U R E F O R
P R O P E R T I E S OF M I X T U R E S
Petroleum fluids are mixtures of hydrocarbon compounds,
which in the reservoirs or during processing could he in the
form of liquid, gas, or vapor. Some heavy products such as
asphalts and waxes are in solid forms. But in petroleum proc-
essing most products are in the form of liquid under atmo-
spheric conditions. The same liquid products during proc-
essing might be in a vapor form before they are stored as a
product. Certain properties such as critical constants, acen-
tric factor, and molecular weight are specifications of a com-
pound regardless of being vapor of liquid. However, physical
properties such as density, transport, or thermal properties
depend on the state of the system and in many cases sepa-
rate methods are used to estimate properties of liquid and
gases as will be discussed in the following chapters. In this
section a general approach toward calculation of such prop-
erties for liquids and gases with known compositions is pre-
sented. Since density and refractive index are important phys-
ical properties in characterization or petroleum fractions they
are used in this section to demonstrate our approach for mix-
ture properties. The same approach will be applied to other
properties throughout the book.
3.4.1 L i q u i d M i x t u r e s
In liquid systems the distance between molecules is much
smaller than in the case of gases and for this reason the inter-
action between molecules is stronger in liquids. Therefore, the
knowledge of types of molecules in the liquid mixtures is more
desirable than in gas mixtures, especially when the mixture
constituents differ significantly in size and type. For example,
consider two liquid mixtures, one a mixture of a paraffinic hy-
drocarbon such as n-eicosane (n-C20) with an aromatic com-
pound such as benzene (C6) and the second one a mixture of
benzene and toluene, which are both aromatic with close
molecular weight and size. The role of composition in the
n-C20-benzene mixture is much more important than the
role of composition in the benzene-toluene mixture. Simi-
larly the role of type of composition (weight, mole, or volume
fraction) is more effective in mixtures of dissimilar con-
stituents than mixtures of similar compounds. It is for this
reason that for narrow-range petroleum fractions, use of the
119
PNA composition in terms of weight, volume, or mole does
not seriously affect the predicted mixture properties. Use of
bulk properties such as Tb and SG to calculate mixture prop-
erties as described for petroleum fractions cannot be used for
a synthetic and ternary mixture of C5-C10-C25. Another exam-
ple of a mixture that bulk properties directly cannot be used to
calculate its properties is a crude oil or a reservoir fluid. For
such mixtures exact knowledge of composition is required
and based on an appropriate mixing rule a certain physical
property for the mixture may be estimated. The most simple
and practical mixing rule that is applicable to most physical
properties is as follows:
N
(3.45) 0m = E XiOi
i=1
where xi is the fraction of component i in the mixture, Oi is a
property for pure component i, and 0m is property of the mix-
ture with N component. This mixing rule is known as Kay
mixing rule after W. B. Kay of Ohio State, who studied mix-
ture properties, especially the pseudocritical properties in the
1930s and following several decades. Other forms of mixing
rules for critical constants will be discussed in Chapter 5 and
more accurate methods of calculation of mixture properties
are presented in Chapter 6.
Equation (3.45) can be applied to any property such as crit-
ical properties, molecular weight, density, refractive index,
heat capacity, etc. There are various modified version of Eq.
(3.45) when it is applied to different properties. Type of com-
position used forxi depends on the type of property. For exam-
ple, to calculate molecular weight of the mixture (0 = M) the
most appropriate type of composition is mole fraction. Sim-
ilarly mole fraction is used for many other properties such
as critical properties, acentric factor, and molar properties
(i.e., molar heat capacity). However, when Eq. (3.45) is ap-
plied to density, specific gravity, or refractive index parameter
[I = (n2 - 1)/(n 2 + 2)], volume fraction should be used for xi.
For these properties the following mixing rule may also be
applied instead of Eq. (3.45) if weight fraction is used:
N
(3.46) 1/0m = EXwi/Oi
i=l
where Xwi is the weight fraction and the equation can be ap-
plied to d, SG, or parameter I. In calculation of these proper-
ties for a mixture, using Eq. (3.45) with volume fraction and
Eq. (3.46) with weight fraction gives similar results. Applica-
tion of these equations in calculation of mixture properties
will be demonstrated in the next chapter to calculate proper-
ties of crude oils and reservoir fluids.
For liquid mixtures the mixing rule should be applied to the
final desired property rather than to the input parameters. For
example, a property such as viscosity is calculated through
a generalized correlation that requires critical properties as
the input parameters. Equation (3.45) may be applied first
to calculate mixture pseudocritical properties and then mix-
ture viscosity is calculated from the generalized correlation.
An alternative approach is to calculate viscosity of individual
components in the mixture for the generalized correlation
and then the mixing rule is directly applied to viscosity. As
it is shown in the following chapters the second approach
gives more accurate results for properties of liquid mixtures,
120 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

while for gaseous mixtures there is no significant difference where Mg is the gas molecular weight. Density of both a gas
between these two methods. mixture and air at STP can be calculated from Eq. (3.47).
Mgas Psc
3.4.2 Gas Mixtures (3.48) SGg - Pga~ _ 83.14Ts~ _ Mg~
Pair M~irPsc Mair
As discussed earlier the gases at atmospheric pressure condi-
83.14Tsc
tion have much larger free space between molecules than do
liquids. As a result the interaction between various like and where sc indicates the standard condition. Molecular weight
unlike molecules in a gaseous state is less than the molecular of air can be calculated from Eq. (3.48) with molecular
interactions in similar liquid mixtures. Therefore, the role of weight of its constituents obtained from Table 2.1 as MN2 =
composition on properties of gas mixtures is not as strong as 28.01, 3/lo2 = 32.00, and MAr = 39.94. With composition given
in the case of liquids. Of course the effect of composition on as YN2 = 0.78, Yo2 = 0.21, and YA~= 0.01, from Eq. (3.1) we
properties of gas mixtures increases as pressure increases and get Mair = 28.97 g/mol. Equation (2.6) can be derived from
free space between molecules decreases. The role of compo- substituting this value for Mair in Eq. (3.49). In practical cal-
sition on properties of dense gases cannot be ignored. Under culations molecular weight of air is rounded to 29. If for a gas
low-pressure conditions where most gases behave like ideal mixture, specific gravity is known its molecular weight can be
gases all gas mixtures regardless of their composition have the calculated as
same molar density at the same temperature and pressure. As
it will be discussed in Chapter 5, at the standard conditions (3.49) Mg = 29SGg
(SC) of 1.01325 bar and 298 K (14.7 psia and 60~ most gases
where SGg is the gas specific gravity. It should be noted that
behave like ideal gas and RT/P represents the molar volume
values of specific gravity given for certain gases in Table 2.1
of a pure gas or a gas mixture. However, the absolute density
are relative to density of water for a liquefied gas and are
varies from one gas to another as following:
different in definition with gas specific gravity defined from
Mmix P Eq. (2.6). #
(3.47) PmLx -- - -
83.14T
where Pmix is the absolute density of gas mixture in g/cm 3,
3.5 P R E D I C T I O N OF T H E C O M P O S I T I O N
Mmi~ is the molecular weight of the mixture in g/tool, P is
OF P E T R O L E U M F R A C T I O N S
pressure in bar, and T is the temperature in kelvin. Equation
(3.1) can be used to calculate molecular weight of a gas mix-
As discussed earlier the quality and properties of a petroleum
ture, MmL~.However, the mole fraction of component i in a
fraction or a petroleum product depend mainly on the
gas mixture is usually shown as Yi to distinguish from com-
mixture composition. As experimental measurement of the
position of liquid mixtures designated by x~. From definition
composition is time-consuming and costly the predictive
of mole and volume fractions in Section 1.7.15 and use of
methods play an important role in determining the quality
Eq. (3.47) it can be shown that for ideal gas mixtures the
of a petroleum product. In addition the pseudocomponent
mole and volume fractions are identical. Generally volume
method to predict properties of a petroleum fraction requires
and mole fractions are used interchangeably for all types of
the knowledge of PNA composition. Exact prediction of all
gas mixtures. Composition of gas mixtures is rarely expressed
components available in a petroleum mixture is nearly im-
in terms of weight fraction and this type of composition has
possible. In fact there are very few methods available in the
very limited application for gas systems. Whenever composi-
literature that are used to predict the composition. These
tion in a gas mixture is expressed only in percentage it should
methods are mainly capable of predicting the amounts (in
be considered as tool% or vol%. Gas mixtures that are mainly
percentages) of paraffins, naphthenes, and aromatic as the
composed of very few components, such as natural gases, it is
main hydrocarbon groups in all types of petroleum fractions.
possible to consider them as a single pseudocomponent and
These methods assume that the mixture is free of olefinic hy-
to predict properties form specific gravity as the sole param-
drocarbons, which is true for most fractions and petroleum
eter available. This method of predicting properties of nat-
products as olefins are unstable compounds. In addition to the
ural gases is presented in Chapter 4 where characterization
PNA composition, elemental composition provides some vital
of reservoir fluids is discussed. The following example shows
information on the quality of a petroleum fraction or crude
derivation of the relation between gas phase specific gravity
oil. Quality of a fuel is directly related to the hydrogen and
and molecular weight of gas mixture.
sulfur contents. A fuel with higher hydrogen or lower carbon
content is more valuable and has higher heating value. High
Example 3.14 Specific gravity of gases is defined as the ra- sulfur content fuels and crude oils require more processing
tio of density of gas to density of dry air both measured at cost and are less valuable and desirable. Methods of predict-
the standard temperature and pressure (STP). Composition ing amounts of C, H, and S% are presented in the following
of dry air in tool% is 78% nitrogen, 21% oxygen, and 1% ar- section.
gon. Derive Eq. (2.6) for the specific gravity of a gas mixture.

Solution--Equation (2.6) gives the gas specific gravity as 3.5.1 Prediction o f PNA Composition
Parameters that are capable of identifying hydrocarbon types
(2.6) SGg- Mg are called characterization parameters. The best example of
28.97
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS
such a parameter is the Watson characterization factor, which
along with other parameters is introduced and discussed in
this section. However, the first known method to predict the
PNA composition is the n-d-M method proposed by Van Nes
and Van Westen [30] in the 1950s. The n-d-M method is
also included in the ASTM manual under ASTM D 3238 test
method. The main limitation of this method is that it can-
not be applied to light fractions. Later in the 1980s Riazi and
Daubert [36, 47] proposed a series of correlations based on
careful analysis of various characterization parameters. The
unique feature of these correlations is that they are appli-
cable to both light and heavy fractions and identify various
types of aromatics in the mixture. In addition various meth-
ods are proposed based on different bulk properties of the
mixture that might be available. The Riazi-Daubert methods
have been adopted by the API Committee on characteriza-
tion of petroleum fractions and are included in the fourth
and subsequent editions of the API-TDB [2] since the early
1980s. The other method that is reported in some literature
sources is the Bergamn's method developed in the 1970s [48].
This method is based on the Watson K and specific gravity
of the fraction as two main characterization parameters. One
c o m m o n deficiency for all of these methods is that they do
not identify n-paraffins and isoparaffins from each other. In
fact compositional types of PIONA, PONA, and PINA can-
not be determined from any of the methods available in the
literature. These methods provide m i n i m u m information on
the composition that is predictive of the PNA content. This is
mainly due to the complexity of petroleum mixtures and dif-
ficulty of predicting the composition from measurable bulk
properties. The method of Riazi-Daubert, however, is capable
of predicting the monoaromatic (MA) and polyaromatic (PA)
content of petroleum fractions.
In general low boiling point fractions have higher paraffinic
and lower aromatic contents while as boiling point of the frac-
tion increases the amount of aromatic content also increases.
In the direction of increase in boiling point, in addition to
FIG. 3.22mVariation of composition of petroleum fractions with boiling point. Reprinted
from Ref. [7], p. 469, by courtesy of Marcel Dekker, Inc.
121
aromatic content, amounts of sulfur, nitrogen, and other het-
eroatoms also increase as shown in Fig. 3.22.
3.5.1.1 Characterization Parameters for Molecular
Type Analysis
A characterization parameter that is useful for molecular type
prediction purposes should vary significantly from one hy-
drocarbon type to another. In addition, its range of varia-
tion within a single hydrocarbon family should be minimal.
With such specifications an ideal parameter for character-
izing mo]ecular type should have a constant value within a
single family but different values in different families. Some
of these characterization parameters (i.e., SG, I, VGC, CH,
and Kw), which are useful for molecular type analysis, have
been introduced and defined in Section 2. i. As shown in Table
2.4, specific gravity is a parameter that varies with chemical
structure particularly from one hydrocarbon family to an-
other. Since it also varies within a single family, it is not a
perfect characterizing parameter for molecular type analysis
but it is more suitable than boiling point that varies within a
single family but its variation from one family to another is
not significant. One of the earliest parameters to characterize
hydrocarbon molecular type was defined by Hill and Coats in
1928 [49], who derived an empirical relation between viscos-
ity and specific gravity in terms of viscosity gravity constant
(VGC), which is defined by Eq. (2.15) in Section 2.1.17. Def-
inition of VGC by Eqs. (2.15) or (2.16) limits its application
to viscous oils or fractions with kinematic viscosity at 38~
(100~ above 38 SUS (~3.8 cSt.). For quick hand estima-
tion of VGC from viscosity and specific gravity, ASTM [4] has
provided a homograph, shown in Fig. 3.23, that gives VGC
values close to those calculated from Eq. (2.15). Paraffinic
oils have low VGC, while napthenic oils have high VGC val-
ues. Watson K defined by Eq. (2.13) in terms of MeABP and
SG was originally introduced to identify hydrocarbon type
[9, 50, 51], but as is shown later, this is not a very suitable
parameter to indicate composition of petroleum fractions.
122 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

--41
4.6 is defined by Eq. (2.36). Paraffinic oils have low I values while
I
aromatics have high I values. Carbon-to-hydrogen weight
42 ratio defined in Section 2.1.18 is also a useful parameter
5- ---'13 that indicates degree of hydrocarbon saturation and its value
increases from paraffinic to naphthenic and aromatic oils.
--44-
---45 L02O
Methods of prediction of CH was discussed in Section 2.6.3.
6 - 1.02 Application of the hydrogen-to-carbon ratio in characteriza-
tion of different types of petroleum products is demonstrated
. LOoo .~
7 -7-5o by Fryback [54]. An extensive analysis and comparison of
6 1.00 various characterization parameters useful for prediction of
.980 the composition of petroleum fractions is presented by Ri-
azi and Daubert [36,47]. Comparison of parameters P~, VGC,
IO , 60 .$8
Kw, and I is presented in Table 3.21 and Fig. 3.24. From this
-70 ,960 1 analysis it is clear that parameters/~ and VHC best separate
-940 .96
hydrocarbon types, while parameters Kw and I show large
45, -ao variations for aromatic and naphthenic compounds making
them less suitable for prediction of composition of petroleum
fractions.
8 - o~ ~_ .goo 20 Another very useful parameter that not only separates
paraffins and aromatics but also identifies various hydrocar-
- o 8 89o.1 _~ .gz ,-" bon types is defined through molecular weight and refractive
30- rC3 .~ >
,
..
P-

index as [36]:
.880=- Eo: 25 i <
40- CO <
-200 _~ .~ =~-.90 ,-r
0
(3.50) m = M ( n - 1.475)
50 -840 ~ ~
CO
-_ -zso where n is the refractive index at 20~ Parameter m was de-
60 - aco . ,e2o -: " ~ .as fined based on the observation that refractive index varies lin-
70 ~_: 50
early with 1/M with slope of m for each hydrocarbon group
8 0 -~ ,800
9 0 -- 4tO0 [55]. Values of parameter m for different hydrocarbon groups
I00 -~ - -780 - .86 calculated from Eq. (3.50) are given in Table 3.22.
120 ~ 500 - As shown in Table 3.22, paraffins have low m values while
55
~. - 600 ,;"60 --
140 aromatics have high m values. In addition, paraffinic and
-- - 700
-- ~ _ ~ .84
160 ~ -740 naphthenic oils have negative m values while aromatic oils
180 .-800
zoo --9oo have positive m values. Parameter m nicely identifies vari-
L I000 .720 40
ous aromatic types and its value increases as the number of
- .82
-- 1200 .ZOO tings increases in an aromatic compound. A pure hydrocar-
bon whose m value is calculated as - 9 has to be paraffinic,
500 "-- t 4 0 0
it cannot be naphthenic or aromatic. This parameter is par-
i- 1600 ticularly useful in characterizing various aromatic types in
400 "-- 1800
aromatic-rich fractions such as coal liquids or heavy residues.
FIG. 3 . 2 3 ~ E s t i m a t i o n of V G C from kinematic viscosity and Besides the parameters introduced above there are a num-
specific gravity [4]. ber of other parameters that have been defined for the purpose
of characterizing hydrocarbon type. Among these parameters
Aromatic oils have low Kw values while paraffinic oils have viscosity index (VI) and correlation index (CI) are worth defin-
high Kw values. Kurtz and Ward in 1935 [52] defined refrac- ing. The viscosity index was introduced in 1929 by Dean and
tivity intercept, Ri, in terms of refractive index (n) and density Davis and uses the variation of viscosity with temperature
(d) at 20~ which is presented by Eq. (2.14). The definition as an indication of composition of viscous fractions. It is an
is based on this observation that a plot of refractive index empirical number indicating variation of viscosity of oil with
against density for any homologous hydrocarbon group is lin- temperature. A low VI value indicates large variation of vis-
ear. Ri is high for aromatics and low for naphthenic stocks. cosity with temperature that is a characteristic of aromatic
The most recent characterization parameter was introduced oils. Similarly, paraffinic hydrocarbons have high VI values.
in 1977 by Huang [53] in terms of refractive index and it The method is described under ASTM D 2270-64 [4] and in

TABLE 3.2t--Values of characterization factors.

Value Range
Hydrocarbontype M Ri VGC K I
Paraffin 337-535 1.048-l.05 0.74-0,75 13.1-1.35 0.26-0.273
Naphthene 248--429 1.03-1.046 0.89-0.94 10.5-13.2 0.278-0.308
Aromatic 180-395 1.07-1.105 0.95-1.13 9.5-12.53 0.298-0.362
Takenwith PermissionfromRef. [47].
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S 123

N I[] I A I
Ri I I I I I I I I
1.02 1.04 1.06 1.08 1.00

[] I---'7-]1 A
VGC I I I I I I I I
0.76 0.84 0.92 1.00 1.08

N I
A I
K I I I I I I
10 11 12 13 14

N I
I A I
I I I I I I I I I I I
0.26 0.28 0.30 0.32 0.34 0.36

FIG. 3.24--Comparison of different characterization factors for prediction

of composition of petroleum fractions (see Table 3.20).

the API-TDB [2]. The VI is defined as [2] in tables i n the s t a n d a r d methods (ASTM D 2270, ISO 2909)
as well as in the API-TDB [2]. However, for viscous oils with
(3.51) VI - L - U 100 a k i n e m a t i c viscosity at 100~ of greater t h a n 70 mm2/s, the
D values of L a n d D are given by the following relationships [5]:
where
L = 0.8353Y 2 + 14.67Y - 216
L -- kinematic viscosity of reference oil at 40~ with 0
VI oil, cSt. (3.52) D = 0.6669Y 2 + 2.82Y - 119
U -- k i n e m a t i c viscosity of oil at 40~ whose VI is to be where Y is the k i n e m a t i c viscosity of oil whose VI is to be cal-
calculated, cSt. culated at 100~ i n cSt. I n English units in which reference
D = L - H, i n cSt. t e m p e r a t u r e s of 37.8 a n d 98.9~ (100 a n d 210~ are used,
H = kinematic viscosity of reference oil at 40 ~C with 100 values of n u m e r i c a l coefficients in Eq. (3.52) are slightly dif-
VI oil, cSt. ferent a n d are given i n the API-TDB [2].
The reference oils with 0 a n d 100 VI a n d the oil whose
VI is to be calculated have the same k i n e m a t i c viscosity at L = 1.01523Y2 + 1 2 . 1 5 4 Y - 155.61
(3.53)
100~ I n English units of system, VI is defined in terms of D ----0.8236Y 2 - 0.5015Y - 53.03
viscosity at 37.8 a n d 98.9~ which correspond to 100 a n d
210~ respectively [2]. However, in the SI units, viscosity at Viscosity index defined by Dean a n d Davis in the form of
reference t e m p e r a t u r e s of 40 a n d 100~ have b e e n used to Eq. (3.51) does not work very well for oils with VI values of
define the VI [5]. For oils with a kinematic viscosity of 100~ greater t h a n 100. For such oils ASTM D 2270 describes the
of less t h a n 70 mm2/s (cSt.), the values of L a n d D are given calculation m e t h o d a n d it is s u m m a r i z e d below for viscosity
at reference t e m p e r a t u r e s of 40 a n d 100~ [4, 5]:
10 N -- 1
TABLE 3.22--Values of parameter m for different types (3.54) VI- - - + 100
of hydrocarbons [36]. 0.00715
Hydrocarbontype m where N is given by the following relation:
Paraffins -8.79
Cyclopentanes - 5.41 (3.55) N = log H - log U
Cyclohexanes -4.43 logY
Benzenes 2.64
Naphthalenes 19.5 i n which U a n d Y are k i n e m a t i c viscosity of oil i n cSt whose
Condensed Tricyclics 43.6 VI is to be calculated at 40 a n d 100~ respectively. Values of
 124 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

TABLE 3.23---Comparison of VI, VGC, Kw, and CI for several groups of oils [29].
Type of oil VI VGC Kw CI
High VI distillate 100 0.800-0.805 12.2-12.5 < 15
Medium VI distillate 80 0.830-0.840 11.8-12.0 15-50
Low VI distillate 0 0.865-0.890 11.0-11.5 >50
Solvent Extracts ... 0.880-0.950 10.0-11.0 ...
Recycle Stock ... 0.900-0.950 10.0-11.0 ...
Cracked Residues ... 0.950-1.000 9.8-11.0 ...

H versus k i n e m a t i c viscosity of oil at 100 ~ (Y) w h e n Y is less
t h a n 70 cSt are t a b u l a t e d in the ASTM D 2270 m e t h o d [4]. F o r
oils with values of Y greater t h a n 70 cSt H m u s t be c a l c u l a t e d
from the relation given b e l o w [5].
(3.56) H = 0.1684Y 2 + 1 1 . 8 5 Y - 97
If kinematic viscosities are available in English units at 37.8
a n d 98.9~ (100 a n d 210~ t h e n Eqs. (3.54) a n d (3.56) should
be r e p l a c e d with the following relations as given in the API-
TDB [2]:
10 N - I
(3.57) VI- - - + 100
0.0075
(3.58) H = 0.19176Y 2 + 1 2 . 6 5 5 9 Y - 102.58
in w h i c h Eq. (3.58) m u s t be u s e d for fractions with k i n e m a t i c
viscosity at 99~ (210~ greater t h a n 75 cSt [2].
Finally the correlation index (CI) defined b y the U.S. B u r e a u
of Mines is expressed b y the following e q u a t i o n [7]:
48640
(3.59) CI - + 473.7SG - 456.8
rb
in w h i c h Tb is the v o l u m e average boiling p o i n t (VABP) in
kelvin. Values of CI b e t w e e n 0 a n d 15 indicate a p r e d o m i -
n a n t l y paraffinic oil. A value of CI greater t h a n 50 indicates a
p r e d o m i n a n c e of a r o m a t i c c o m p o u n d s [7]. It has a t e n d e n c y
to increase with increasing boiling p o i n t in a given c r u d e oil. A
c o m p a r i s o n b e t w e e n values ofVI, VGC, Kw, a n d CI for several
types of p e t r o l e u m fractions a n d p r o d u c t s is p r e s e n t e d in Ta-
ble 3.23. A c o m p l e t e c o m p a r i s o n of various c h a r a c t e r i z a t i o n
p a r a m e t e r s i n d i c a t i n g c o m p o s i t i o n of p e t r o l e u m fractions for
the three h y d r o c a r b o n groups is p r e s e n t e d in Table 3.24. All
p a r a m e t e r s except Ri, Kw, a n d VI increase in the direction
from paraffinic to n a p h t h e n i c a n d a r o m a t i c oils.
Example 3.15--Calculate viscosity index of an oil having
k i n e m a t i c viscosities of 1000 a n d 100 cSt at 37.8 a n d 98.9~
(100 a n d 210~ respectively.
Solution--For this oil 1)99(210) > 70 cSt, thus we can use Eqs.
(3.51)-(3.57) for calculation of VI. Since the VI is n o t k n o w n
we a s s u m e it is greater t h a n 100 a n d we use Eqs. (3.53)-(3.57)
to calculate the VI. However, since k i n e m a t i c viscosities are
given at 38 a n d 99~ Eqs. (3.57) a n d (3.58) s h o u l d be used.
F r o m the i n f o r m a t i o n given U = i000 a n d Y = 100. Since Y
is greater t h a n 75 cSt, Eq. (3.58) m u s t be used to calculate
p a r a m e t e r H, w h i c h gives H -- 3080.6. F r o m Eqs. (3.55) a n d
(3.57) we get N = 0.2443 a n d VI--- 200.7. Since calculated VI is
g r e a t e r t h a n 100, the initial a s s u m p t i o n is correct. Otherwise,
Eq. (3.51) m u s t be used. Since the value of VI is quite high
the oil is paraffinic as s h o w n in Table 3.24. If i)99(210) was less
t h a n 70 cSt then tables p r o v i d e d by ASTM [4] o r API-TDB [2]
should be used to calculate p a r a m e t e r s L a n d D.
3.5.1.2 A P I (Riazi-Daubert) Methods
To develop a m e t h o d for p r e d i c t i n g the c o m p o s i t i o n of olefin-
free p e t r o l e u m fractions three equations are r e q u i r e d to ob-
tain fractions of paraffins (xp), n a p h t h e n e s (XN), a n d a r o m a t -
ics (XA). The first a n d m o s t obvious e q u a t i o n is k n o w n f r o m
the m a t e r i a l balance:
(3.60) Xp + XN + XA = 1
Two o t h e r equations can be established by a p p l y i n g Eq. (3.40)
for two p a r a m e t e r s that can c h a r a c t e r i z e h y d r o c a r b o n types
a n d are easily m e a s u r a b l e . Analysis of various characteriza-
tion factors s h o w n in Table 3.21 a n d Fig. 3.24 indicates t h a t / ~
a n d VGC are the m o s t suitable p a r a m e t e r s to identify hydro-
c a r b o n type. F o r example, if for a p u r e h y d r o c a r b o n / ~ = 1.04,
it has to be a n a p h t h e n i c h y d r o c a r b o n , it c a n n o t be paraffinic
o r a r o m a t i c since only for the n a p h t h e n i c g r o u p /~ varies
f r o m 1.03 to 1.046. Refractivity intercept has been related
to the p e r c e n t o f n a p h t h e n i c c a r b o n a t o m s (%CN) as /~ =
1.05-0.0002 %CN [7]. Riazi a n d D a u b e r t [47] u s e d b o t h / ~
a n d VGC to develop a predictive m e t h o d for the c o m p o s i t i o n
of viscous p e t r o l e u m fractions. Properties of p u r e hydrocar-
b o n s from the API RP-42 [56] have b e e n u s e d to c a l c u l a t e / ~
a n d VGC for a n u m b e r of heavy h y d r o c a r b o n s with molecu-
lar weights greater t h a n 200 as shown in Table 2.3 a n d Table
3.21. Based on the values o f / ~ a n d VGC for all h y d r o c a r b o n s ,
average values of these p a r a m e t e r s were d e t e r m i n e d for the
three g r o u p s of paraffins, n a p h t h e n e s , a n d aromatics. These
average values f o r / ~ are as follows: 1.0482 (P), 1.0138 (N),
a n d 1.081 (A). S i m i l a r average values for VGC are 0.744 (P),

TABLE 3.24--Comparison of various characterization parameters for molecular type analysis.

Parameter Defined by Eq. (s) Paraffins Naphthenes Aromatics
/~ (2.14) medium low high
VGC (2.15) or (2.16) low medium high
m (3.50) low medium high
SG (2.2) low medium high
I (2.3) (2.36) low medium high
CI (3.59) low medium high
Kw (2.13) high medium low
VI (3.51)-(3.58) high medium low
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S
0.915 (N), a n d 1.04 (A). Applying Eq. (3.40) for Ri a n d VGC
gives the following two relations.
(3.61) /~ = 1.0482xp + 1.038XN + 1.081XA
(3.62) VGC = 0.744Xp + 0.915XN + 1.04XA
A regression of 33 defined h y d r o c a r b o n mixtures changes
the n u m e r i c a l constants in the above equations b y less t h a n
2% as follows [29, 47]:
(3.63) Ri = 1.0486xv + 1.022XN + 1.1 lXA
(3.64) VGC = 0.7426xv + 0.9XN + 1.112XA
S i m u l t a n e o u s solution of Eqs. (3.60), (3.63), a n d (3.64)
gives the following equations for e s t i m a t i o n of the PNA com-
p o s i t i o n of fractions with M > 200.
(3.65) Xp = - 9 . 0 + 12.53/~ - 4.228VGC
(3.66) xN = 18.66 - 19.9/~ + 2.973VGC
(3.67) XA = --8.66 + 7.37/~ + 1.255VGC
These equations c a n be a p p l i e d to fractions with m o l e c u l a r
weights in the r a n g e of 200-600. As m e n t i o n e d earlier, Xp,
Xn, a n d XA calculated f r o m the above relations m a y r e p r e s e n t
volume, mole, or weight fractions. E q u a t i o n s (3.65)-(3.67)
c a n n o t be a p p l i e d to light fractions having k i n e m a t i c viscosity
at 38~ of less t h a n 38 SUS (~3.6 cSt.). This is b e c a u s e VGC
c a n n o t be d e t e r m i n e d as defined by Eqs. (2.15) a n d (2.16). F o r
such fractions Riazi a n d D a u b e r t [47] defined a p a r a m e t e r
s i m i l a r to VGC called viscosity gravity function, VGF, b y the
following relations:
(3.68) VGF = - 1.816 + 3.484SG - 0.1156 In v38(100)
(3.69) VGF = - 1 . 9 4 8 + 3.535SG - 0.1613 In i)99(210)
w h e r e 1)38(100) a n d 1)99(210) are k i n e m a t i c viscosity in mm2/s
(cSt) at 38 a n d 99~ (100 a n d 210~ respectively. F o r a
p e t r o l e u m fraction, b o t h Eqs. (3.68) a n d (3.69) give nearly the
s a m e value for VGF; however, if k i n e m a t i c viscosity at 38~ is
available Eq. (3.68) is preferable for calculation of VGE These
equations have b e e n derived b a s e d on the observation t h a t at
a fixed t e m p e r a t u r e , plot of SG versus In 1) is a linear line for
each h o m o l o g o u s h y d r o c a r b o n group, b u t each group has
a specific slope. F u r t h e r i n f o r m a t i o n on derivation of these
equations is p r o v i d e d b y Riazi a n d D a u b e r t [47]. P a r a m e t e r
VGF is basically defined for fractions with m o l e c u l a r weights
of less t h a n 200. Based on the c o m p o s i t i o n of 45 defined
mixtures (synthetic) a n d w i t h a n a p p r o a c h s i m i l a r to the
one u s e d to develop Eqs. (3.65)-(3.67), three relationships in
t e r m s o f / ~ a n d VGF have b e e n o b t a i n e d to e s t i m a t e the PNA
c o m p o s i t i o n (Xp, XN, XA) of light (M < 200) fractions [47].
These equations were later modified with a d d i t i o n a l d a t a for
b o t h light a n d heavy fractions a n d are given b e l o w [36].
F o r fractions with M <_ 200
(3.70) xv -- - 1 3 . 3 5 9 + 14.4591R~ - 1.41344VGF
(3.71) xn = 23.9825 - 23.33304Ri + 0.81517VGF
(3.72) Xg = 1 -- (Xp + XN)
F o r fractions with M > 200
(3.73) Xv = 2.5737 + 1.0133/~ - 3.573VGC
(3.74) Xs = 2.464 -- 3.6701/~ + 1.96312VGC
125
In these set of equations XA m u s t be calculated from Eq.
(3.72). F o r cases t h a t calculated XA is negative it s h o u l d be
set equal to zero a n d values of Xp a n d XN m u s t be n o r m a l i z e d
in a w a y t h a t Xp + xr~ = 1. The s a m e p r o c e d u r e s h o u l d be
a p p l i e d to Xp o r Xn if one of t h e m calculated from the
above equations is negative. F o r 85 s a m p l e s Eqs. (3.70) a n d
(3.72) give average deviation of 0.04 a n d 0.06 for xv a n d
XN, respectively. F o r 72 heavy fractions, Eqs. (3.72)-(3.74)
p r e d i c t Xp, XN, a n d XA w i t h average deviations of 0.03, 0.04,
a n d 0.02, respectively [36]. These deviations are w i t h i n the
range of e x p e r i m e n t a l u n c e r t a i n t y for the PNA c o m p o s i t i o n .
E q u a t i o n s (3.70)-(3.74) are r e c o m m e n d e d to be u s e d if
e x p e r i m e n t a l d a t a on viscosity are available. F o r cases t h a t
n2o a n d d2o are not available, they can be accurately e s t i m a t e d
f r o m the m e t h o d s p r e s e n t e d in C h a p t e r 2.
F o r fractions t h a t k i n e m a t i c viscosity is n o t available, Riazi
a n d D a u b e r t [36] developed a series of correlations in t e r m s of
o t h e r c h a r a c t e r i z a t i o n p a r a m e t e r s w h i c h are r e a d i l y available
or predictable. These p a r a m e t e r s are SG, m, a n d CH a n d the
predictive equations for PNA c o m p o s i t i o n are as follows:
F o r fractions w i t h M < 200
(3.75) Xp = 2.57 - 2.877SG + 0.02876CH
(3.76) XN = 0.52641 -- 0.7494Xp -- 0.021811m
or
(3.77) Xp = 3.7387 - 4.0829SG + 0.014772m
(3.78) XN = - 1 . 5 0 2 7 + 2.10152SG - 0.02388m
F o r fractions with M > 200
(3.79) xp = 1.9842 - 0.27722Ri - 0.15643CH
(3.80) XN = 0.5977 -- 0.761745Ri + 0.068048CH
or
(3.81) xv -- 1.9382 + 0 . 0 7 4 8 5 5 m - 0.19966CH
(3.82) XN -------0.4226 -- 0 . 0 0 7 7 7 m + 0.107625CH
I n all of these cases XA m u s t be calculated f r o m Eq. (3.72).
E q u a t i o n s (3.75) a n d (3.76) have b e e n evaluated with PNA
c o m p o s i t i o n of 85 fractions in the m o l e c u l a r weight range
of 78-214 a n d give average deviations of 0.05, 0.08, a n d
0.07 for xv, XN, a n d xA, respectively. F o r the s a m e d a t a set
Eqs. (3.77) a n d (3.78) give AAD of 0.05, 0.086, a n d 0.055 for
Xp, XN, a n d xa, respectively. F o r 72 fractions w i t h molecu-
lar weight range of 230-570, Eqs. (3.79)-(3.82) give n e a r l y
the s a m e AAD of 0.06, 0.06, a n d 0.02 for Xp, xN, a n d XA,
respectively. I n cases t h a t i n p u t p a r a m e t e r s for the above
m e t h o d s are n o t available Eqs. (3.77) a n d (3.78) in t e r m s of
SG a n d m are m o r e suitable t h a n o t h e r equations since re-
fractive index a n d m o l e c u l a r weight can be e s t i m a t e d m o r e
a c c u r a t e l y t h a n CH. Although Eqs. (3.77) a n d (3.78) have
been derived from a d a t a set on fractions w i t h m o l e c u l a r
weights u p to 200, they can be safely u s e d u p to m o l e c u -
lar weight of 300 w i t h o u t serious errors. Most recently, Eqs.
(3.77) a n d (3.78) have b e e n modified to e x p a n d the range
of a p p l i c a t i o n of these equations for heavier fractions, b u t
in general their a c c u r a c y is n o t significantly different from
the equations p r e s e n t e d here [45]. F o r example, for frac-
tions w i t h 70 < M < 250, Riazi a n d R o o m i [45] modified Eqs.
126 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

(3.77) and (3.78) as: Xp -- 3 . 2 5 7 4 - 3.48148 SG + 0.011666m get XMA= --0.228 and XpA= -0.025. Since both numbers are
andxN = --1.9571 + 2.63853 SG - 0.03992m. Most of the data negative the actual estimated values are XMA= 0 and XpA= 0.
used in development of Eqs. (3.70)-(3.82) were in terms of From Eq. (3.85), ;cA = 0, which is consistent with the previous
volume fractions for xp, XN, and Xg. Therefore, generally esti- result from Eqs. (3.77), (3.78), and (3.72). t
mated values represent volume fractions; however, they can
be used as mole fractions without serious errors. 3.5.1.3 A P I M e t h o d
In all the above equations total aromatic content is cal- Since 1982 API has adopted the methods developed by Ri-
culated from Eq. (3.72). As discussed earlier for cases that azi and Daubert [36, 47] for prediction of the composition
aromatic content is high it should be split into two parts for of petroleum fractions. Equations (3.65)-(3.67) and similar
a more accurate representation of hydrocarbon types in a equations developed for light fractions in terms of Ri and
petroleum mixture. Aromatics are divided into monoaromat- VGF by Riazi and Daubert in 1980 [47] were included in
ics (XMA)and polyaromatics (XPA) and the following relations the fourth edition of the API-TDB-82. However, after devel-
have been derived for fractions with molecular weights of less opment of Eqs. (3.70)-(3.74) in terms of viscosity and Eqs.
than 250 [36]: (3.83)-(3.84) for prediction of the amount of different types
(3.83) XMA= --62.8245 + 59.90816R~ - 0.0248335m of aromatics in 1986 [36], they were included in the fifth and
subsequent editions of the API-TDB [2]. The API methods for
(3.84) XpA= 11.88175 -- 11.2213R~ + 0.023745m prediction of the composition of petroleum fractions require
(3.85) XA -~- XMA -~- XpA kinematic viscosity at 38 or 99~ and if not available, it should
be estimated from Eq. (2.128) or (2.129) in Chapter 2.
The equations may be applied to fractions with total aromatic
content in the range of 0.05-0.96 and molecular weight range 3.5.1.4 n-doM M e t h o d
of 80-250. Based on a data set for aromatic contents of 75
This method requires three physical properties of refractive
coal liquid sample, Eqs. (3.83)-(3.85) give AAD of about 0.055,
index (n20), density (d20), and molecular weight (M). For this
0.065, and 0.063 for XMA,XPA,and XA, respectively. Maximum
reason the method is called n-d-M method and is the oldest
AD is about 0.24 for Xpg. Equations (3.83) and (3.84) have not
method for prediction of the composition of petroleum frac-
been evaluated against petroleum fractions. For heavier frac-
tions. The method is described in the book by Van Nes and
tions no detailed composition on aromatics of fractions were
Van Westen in 1951 [30] and it is included in the ASTM man-
available; however, if such data become available expressions
ual [4] under ASTM D 3238 test method. The method does
similar to Eqs. (3.83) and (3.84) may be developed for heavier
not directly give the PNA composition, but it calculates the
fractions.
distribution of carbon in paraffins (%Cp), naphthenes (%CN),
and aromatics (%CA). However, since carbon is the dominant
Example 3.16--A gasoline sample produced from an Aus-
element in a petroleum mixture it is assumed that the %Cp,
tralian crude oil has the boiling range of C5-65~ specific
%CN, and %CA distribution is proportional to %P, %N, and %A
gravity of 0.646, and PNA composition of 91, 9, and 0 vol%
distribution. In this assumption the ratio of carbon to hydro-
(Ref. [46], p. 302). Calculate the PNA composition from a suit-
gen is considered constant in various hydrocarbon families.
able method and compare the results with the experimental
Errors caused due to this assumption are within the range of
values.
uncertainty in experimental data reported on the PNA com-
position. Another input data required for this method is sulfur
Solution--For this fraction the only information available
content of the fraction in wt% (%S) and should be known if
are boiling point and specific gravity. From Table 2.1 the boil-
it exceeds 0.206 wt%. The method should not be applied to
ing point of n-C5 is 36~ Therefore, for the fraction the
fractions with sulfur content of greater than 2%. This method
characteristic boiling point is Tb ----(36 + 65)/2 ----50.5~ The
is applicable to fractions with boiling points above gasoline.
other characteristic of the fraction is SG -- 0.646. This is
In addition this method should be applied to fractions with
a light fraction (low Tb) so we use Eq. (2.51) to calculate
ring percent, %CR (%CN + %CA) up to 75% provided that %CA
molecular weight as 84.3. Since viscosity is not known, the
(as determined from the n-d-M method) is not higher than
most suitable method to estimate composition is through
1.5 times %CN [7]. The n-d-M method also provides equations
Eqs. (3.77) and (3.78). They require parameter m, which
for calculation of total number of rings (RT), number of
in turn requires refractive index, n. From Eq. (2.115), I =
aromatic rings (Ra), and number of naphthenic rings (RN)
0.2216 and from Eq. (2.114), n = 1.3616. With use of M and
in an average molecule in the fractions. The method is
n and Eq. (3.50), m = -9.562. From Eq. (3.77) and (3.78),
expressed in two sets of equations, one for n2o, d20 (20~
xp = 0.96 and xN -- 0.083. From Eq. (3.72), ;cA = -4.3. Since
and another set for n70 and d7o (70~ as input data. In this
XA is negative thus ;CA----0 and Xp, XN should be normalized
section correlation in terms of n20 and d20 are presented. The
as xp = 0.96/(0.96 + 0.083) -- 0.92 and XN = l -- 0.92 ----0.08.
other set of correlations for measurement of n and d at 70~
Therefore, the predicted PNA composition is 92, 8, 0% versus
is given in the literature [7, 24].
the experimental values of 91, 9, and 0%.
The aromatic content for this fraction is zero and there %CA = av + 3660/M
is no need to estimate XMA and XpA from Eqs. (3.83) and %CN = %CR -- %CA
(3.84); however, to see the performance of these equations
(3.86) %Cp = 100 - %CR
for this sample we calculate ;CA from Eq. (3.85). From
Eq. (2.113), d = 0.6414 and from Eq. (2.14), R4 -- 1.0409. Us- Rn = 0.44 + bvM
ing these values o f / ~ and m in Eqs. (3.83) and (3.84), we RN = RT-- R,
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS
where
v = 2.51(n - 1.475) - (d - 0.851)
430 ifv > 0
a= 670 if v < 0
b=[0"055 ifv>0
! 0.080 if v < 0
w = ( d - 0.851) - 1.11(n- 1.475)
% C R = [ 8 2 0 w - - 3 % S + 1 0 0 0 0 / M if w > 0
1440w-- 3%S + 10600/M if w < 0
1.33+0.146M(w-0.005x%S) ifw>0
RT= 1 . 3 3 + 0 . 1 8 0 M ( w - 0 . 0 0 5 x % S ) if w < 0
Once carbon distribution is calculated from Eq. (3.86), the
PNA composition can be determined as follows:
xv = %Cv/ l O0
(3.87) XN -----%C~/100
XA= %CA/100
As mentioned above the n-d-M method cannot be applied
to light fractions with molecular weights of less than 200.
However, when it was evaluated against PNA composition
of 70 fractions for the molecular weight range of 230-570,
AAD of 0.064, 0.086, and 0.059 were obtained for xv, XN, and
xh, respectively. Accuracy of the n-d-M method for prediction
of composition of fractions with M > 200 is similar to the
accuracy of Eqs. (3.79)-(3.82). But accuracy of Eqs. (3.73)
and (3.74) in terms of viscosity (API method) is more than
the n-d-M method [30, 36].
In addition to the above methods there are some other
procedures reported in the literature for estimation of the
PNA composition of petroleum fractions. Among these
methods the Bergman's method is included in some refer-
ences [48]. This method calculates the PNA composition in
weight fraction using the boiling point and specific gravity
of the fraction as input data. The weight fraction of aromatic
content is linearly related to Kw. The xv and XN are calculated
through simultaneous solution of Eqs. (3.72) and (3.46)
when they are applied to specific gravity. Specific gravity
of paraffinic, naphthenic, and aromatic pseudocomponents
(SGp, SGN, and SGA) are calculated from boiling point of
the fraction. Equation (2.42) may be used to calculate SG for
different groups from Tb of the fraction. Except in reference
[48] this method is not reported elsewhere. There are some
other specific methods reported in various sources for
each hydrocarbon group. For example, ASTM D 2759 gives a
graphical method to estimate naphthene content of saturated
hydrocarbons (paraffins and naphthenes only) from refrac-
tivity intercept and density at 20~ In some sources aromatic
content of fractions are related to aniline point, hydrogen
content, or to hydrogen-to-carbon (HC) atomic ratio [57]. An
example of these methods is shown in the next section.
3.5.2 Prediction of Elemental Composition
As discussed earlier, knowledge of elemental composition
especially of carbon (%C), hydrogen (%H), and sulfur con-
tent (%S) directly gives information on the quality of a fuel.
Knowledge of hydrogen content of a petroleum fraction helps
to determine the amount of hydrogen needed if it has to go
through a reforming process. Petroleum mixtures with higher
127
hydrogen content or lower carbon content have higher heat-
ing value and contain more saturated hydrocarbons. Predic-
tive methods for such elements are rare and limited so there
is no possibility of comparison of various methods but the
presented procedures are evaluated directly against experi-
mental data.
3.5.2.1 Prediction of Carbon and Hydrogen Contents
The amount of hydrogen content of a petroleum mixture is
directly related to its carbon-to-hydrogen weight ratio, CH.
Higher carbon-to-hydrogen weight ratio is equivalent to lower
hydrogen content. In addition aromatics have lower hydrogen
content than paraffinic compounds and in some references
hydrogen content of a fraction is related to the aromatic con-
tent [57] although such relations are approximate and have
low degrees of accuracy. The reason for such low accuracy
is that the hydrogen content of various types of aromatics
varies with molecular type. Within the aromatic family, dif-
ferent compounds may have different numbers of rings, car-
bon atoms, and hydrogen content. In general more accurate
prediction can be obtained from the CH weight ratio method.
Several methods of estimation of hydrogen and carbon con-
tents are presented here.
3.5.2.1.1 Riazi Method--This method is based on calcula-
tion of CH ratio from the method of Riazi and Daubert given
in Section 2.6.3 and estimation of %S from Riazi method in
Section 3.5.2.2. The main elements in a petroleum fraction
are carbon, hydrogen, and sulfur. Other elements such as ni-
trogen, oxygen, or metals are in such small quantities that
on a wt% basis their presence may be neglected without se-
rious error on the composition of C, H, and S. This is not to
say that the knowledge of the amounts of these elements is
not important but their weight percentages are negligible in
comparison with weight percentages of C, H, and S. Based
on this assumption and from the material balance on these
three main elements we have
(3.88) %C + %H + %S = 100
%C
(3.89) = CH
%H
From simultaneous solution of these two equations, assum-
ing %S is known, the following relations can be obtained for
%H and %C:
100 - %S
(3.90) %H -
1 +CH
(3.91) % C = ( CI~CH ) x ( 1 0 0 - % S )
where %S is the wt% of sulfur in the mixture, which should
be determined from the method presented in Section 3.5.2.2
if the experimental value is not available. Value of CH may
be determined from the methods presented in Section 2.6.3.
In the following methods in which calculation of only %H is
presented, %C can be calculated from Eq. (3.88) if the sulfur
content is available.
3.5.2.1.2 Goossens' Method--Most recently a simple re-
lation was proposed by Goossens to estimate the hydrogen
content of a petroleum fraction based on the assumption of
128 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

molar additivity of structural contributions of carbon types distillation data [4]:

[58]. The correlation is derived from data on 61 oil fractions
with a squared correlation coefficient of 0.999 and average %H = (5.2407 + 0.01448Tb -- 7.018XA)/SG - 0.901XA
deviation of 3% and has the following form: (3.93) + 0.01298XATb -- 0.01345Tb + 5.6879

82.952 - 65.341n
(3.92) %H = 30.346 +
d M
where M is the molecular weight and n and d are refractive
index and density at 20~ respectively. This method should
be applied to fractions with molecular weight range of 84-
459, boiling point range of 60-480~ refractive index range
of 1.38-1.51, and hydrogen content of 12.2-15.6 wt%. In
cases that M is not available it should be estimated from the
Goossens correlation given by Eq. (2.55).
306 where xA is the fraction of aromatics in the mixture and
Tb is an average value of boiling points at 10, 50, and 90
vol% vaporized in kelvin [Tb = (/'10 + Ts0 + %0)/3]. This cor-
relation was developed based on 247 aviation fuels and 84
pure hydrocarbons. This method is quite accurate if all the
input data are available from experimental measurement.
3.5.2.1.4 Jenkins-Walsh M e t h o d - T h e y developed a sim-
ple relation in terms of specific gravity and aniline point in
the following form [59]:

(3.94) %H = 11.17 - 12.89SG + 0.0389AP

3.5.2.1.3 ASTM Method---ASTM describes a method to es-
timate the hydrogen content of aviation fuels under ASTM where AP is the aniline point in kelvin and it may be deter-
D 3343 test method based on the aromatic content and mined from the Winn nomograph (Fig. 2.14) presented in

:1:: 20
...........................................................i.......................................................;................... ...........................................................~..............
.................................... ..................... ............ ........... ........ ~ .................. 7 ..............

6
~iiiiiiiil 4+ 1.27274%H) 1IIIIIIIIIIIII

~0
10

.............................................................. ................... .................... ............ :: 4

, J ; r i
0
5 10 15
(a) H y d r o g e n Weight Percent, % H

.< 70
6O
so
o 40
.2 30

o
20

0
11 12 13 14 15

(b) H y d r o g e n Weight Percent, % H

FIG. 3.25--Relationships between fuel hydrogen content,
(a) CH weight ratio and (b) aromatic content.
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS 129

Section 2.8. The correlation is specifically developed for jet 10

fuels with aniline points in the range of 56-77~ . . . . Venezuelan (5%) . . - "~ ""

There are a number of other methods reported in the liter- 8 ..... Middle East (5%) .-'"
....... West African (5%) .'"
ature. The Winn nomograph may be used to estimate the CH
World Average (5%) . "~
ratio and then %H can be estimated from Eq. (3.90). Fein- . . . . World Average (2%)
9"
..~/,-" . j.s
S"

Wilson-Sherman also related %H to aniline point through 6 : ,- .--

,,,. .j'~ -
API gravity [60]. The oldest and simplest method was pro- ~. : r ~ l"
posed by Bureau of Standards in terms of specific gravity as .." 9 9 ot

given in reference [61 ]:

(3.95) %H = 26 - 15SG
The other simple correlation is derived from data on jet fuels
and is in terms of aromatic content (XA)in the following form
[57]: 0 100 200 300 400 500 600
(3.96) %H = 14.9 - 6.38xg Mid BoilingPoint,~

Finally Fig. 3.25 is based on data from Ref. [57]. Analytical
correlation is also presented in Fig. 3.25(a), which represents
data with an average deviation of 0.5%. Equation (3.96) is pre-
sented in Fig. 3.25(b). When CH ratio is available, %H can be
determined from Fig. 3.25(a) and then %A can be determined
from Fig. 3.25(b).
3.5.2.2 Prediction of Sulfur and Nitrogen Contents
Sulfur is the most important heteroatom that may be present
in a crude oil or petroleum products as dissolved free sulfur
and hydrogen sulfide (H2S). It may also be present as organic
compounds such as thiophenes, mercaptanes, alkyl sulfates,
sulfides (R--S--W), disulfides (R--S--S--R'), or sulfoxides
(R--SO--R'), where R and R' refer to any aliphatic or aromatic
group. Its presence is undesirable for the reasons of corrosion,
catalysts poisoning, bad odor, poor burning, and air pollu-
tion. In addition presence of sulfur in lubricating oils lowers
resistance to oxidation and increases solid deposition on en-
gine parts [62]. New standards and specifications imposed
by governments and environmental authorities in industrial
countries require very low sulfur content in all petroleum
products. For example, reformulated gasolines (RFG) require
sulfur content of less than 300 ppm (<0.03 wt%) [63]. Re-
cently a federal court has upheld an Environmental Protec-
tion Agency (EPA) rule to cut pollution from tractor-trailers
and other large trucks and buses. The rule is expected to re-
duce tailpipe emissions from tractors, buses, and other trucks
up to 90%. The EPA also calls on refineries to reduce the sul-
fur content in diesel oils to 15 ppm by 2007 from the current
level of 500 ppm. The American Lung Association claims that
low-sulfur fuel will reduce the amount of soot from larger
trucks by 90%. This is expected to prevent 8300 premature
deaths, 5500 cases of chronic bronchitis, and another 17600
cases of acute bronchitis in children as provided by the EPA
[64]. Products with high-sulfur contents have low quality and
heating values. Generally, sulfur is associated with heavy and
aromatic compounds [7]. Therefore, high aromatic content
or high boiling point fractions (i.e., residues and coal liquids)
have naturally higher sulfur contents. Distribution of sulfur in
straight-run products of several crude oils and the world aver-
age crude with 2 and 5% sulfur contents is shown in Fig. 3.26.
Data used to generate this figure are taken from Ref. [61]. As
the boiling point of products increases the sulfur content in
the products also increases. However, the distribution of sul-
fur in products may vary from one crude source to another.
FIG. 3.26~Distribution of sulfur in straight-run prod-
ucts for several crude oils. Numbers in the parentheses
indicate sulfur content of crudes.
As boiling point, specific gravity, or aromatic content of
a fraction increases the sulfur content also increases [7].
Parameters Ri, m, and SG have been successfully used to
predict the PNA composition especially aromatic content of
petroleum fractions as shown in Section 3.5.2.1. On this ba-
sis the same parameters have been used for the estimation of
sulfur content of petroleum fractions in the following form
for two ranges of molecular weight.
For fractions with M < 200
(3.97) %S = 177.448 - 170.946Ri + 0.2258m+ 4.054SG
and for fractions with M > 200
(3.98) %S --- -58.02 + 38.463R~ - 0.023m+ 22.4SG
For light fractions in which Eq. (3.96) may give very small
negative values, %S would be considered as zero. Squared cor-
relation coefficients (R2) for these equations are above 0.99.
A summary of evaluation of these equations is presented in
Table 3.25 as given in Ref. [62].
In using these equations parameters n2o, d2o, M, and SG
are required. For samples in which any of these parameters
are not known they can be estimated from the methods dis-
cussed earlier in this chapter. In Chapter 4, it is shown how
this method can be used to estimate sulfur content of whole
crudes. The author is not familiar with any other analytical
method for estimation of sulfur content of petroleum frac-
tions reported in the literature so a comparison with other
methods is not presented. Generally amount of sulfur in var-
ious products is tabulated for various crudes based on the
sulfur content of each crude [61 ].
Another heteroatom whose presence has adverse effect on
the stability of the finished product and processing catalysts
is nitrogen. High nitrogen content fractions require high hy-
drogen consumption in hydro processes. Nitrogen content of
crudes varies from 0.01 to 0.9 wt%. Most of the compounds
having nitrogen have boiling points above 400~ (~750~
and are generally within the aromatic group. Crudes with
higher asphaltene contents have higher nitrogen content as
130 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
TABLE 3.25--Prediction of sulfur content of petroleum fractions [62].
Fraction type No. of point Mol% range SG range Sulfur wt% range
Light 76 76-247 0.57-0.86
Heavy 56 230-1500 0.80-1.05
Overall 132 76-1500 0.57-1.05
~AAD%= Absolute average deviation, %; MAD%= maximum average deviation, %.
well. S i m i l a r to sulfui, n i t r o g e n content of various p e t r o l e u m
fractions is p r e s e n t e d in t e r m s of n i t r o g e n content of the
crude oil [61]. Ball et al. [65] have s h o w n that nitrogen con-
tent of crude oils for each geological p e r i o d is linearly related
to c a r b o n r e s i d u e o f the crude. However, the c o r r e l a t i o n does
not provide i n f o r m a t i o n on n i t r o g e n c o n t e n t of p e t r o l e u m
products. I n general n i t r o g e n c o n t e n t of fractions whose m i d
boiling p o i n t is less t h a n 450~ have n i t r o g e n contents less
t h a n t h a t of crude a n d for heavier cuts the n i t r o g e n wt% in the
fraction is g r e a t e r t h a n that of crude [61]. However, the value
of 450~ at w h i c h nitrogen c o n t e n t of the fraction is n e a r l y the
s a m e as that of crude is a p p r o x i m a t e a n d it m a y vary slightly
with the type of the crude. D a t a r e p o r t e d in Ref. [61] for dis-
t r i b u t i o n o f n i t r o g e n c o n t e n t o f straight r u n distillates have
been c o r r e l a t e d in the following form:
%N2 in fraction
= - 0 . 4 6 3 9 + 8.47267T - 28.9448T 2
%N2 in crude
(3.99) + 27.8155T 3
w h e r e T = Tb/1000 in w h i c h Tb is the m i d boiling p o i n t of the
cut in kelvin. This equation is valid for cuts with m i d boiling
points greater t h a n 220~ a n d is not a p p l i c a b l e to finished
p e t r o l e u m products. A m o u n t of n i t r o g e n in a t m o s p h e r i c
distillates is quite small on p e r c e n t basis. The wt% ratio in
Eq. (3.99) can be r e p l a c e d b y p p m weight ratio for small quan-
tities of nitrogen. E s t i m a t i o n of c o m p o s i t i o n of elements is
d e m o n s t r a t e d in E x a m p l e s 3.17 a n d 3.18.
Example 3 . 1 7 ~ A p e t r o l e u m fraction with a boiling range of
250-300~ is p r o d u c e d from a Venezuelan c r u d e oil (Ref. [46],
p. 360). E x p e r i m e n t a l l y m e a s u r e d p r o p e r t i e s are as follows:
ASTM distillation 262.2,268.3, a n d 278.9~ at 10, 50, a n d 90
vol% recovered, respectively; specific gravity 0.8597; c a r b o n -
to-hydrogen weight ratio 6.69; aniline p o i n t 62~ a r o m a t i c
c o n t e n t 34.9%; a n d sulfur wt% 0.8. E s t i m a t e sulfur c o n t e n t
of the fraction from the m e t h o d p r e s e n t e d in Section 3.5.2.2.
Also calculate %C a n d % H from the following methods: exper-
i m e n t a l data, Riazi, Goossens, ASTM, Jenkins-Watsh, B u r e a u
of Mines a n d Eq. (3.96).
Solution--To e s t i m a t e the sulfur content, p a r a m e t e r s M, n2o,
a n d d20 are r e q u i r e d as the i n p u t data. The fraction is a nar-
r o w fraction a n d the boiling p o i n t at 50% distilled c a n be
c o n s i d e r e d as the characteristic average boiling point, Tb =
268.3~ = 541.5 K. This is a light fraction w i t h M < 300;
therefore, M, d2o, a n d n20 a r e c a l c u l a t e d f r o m Eqs. (2.50) a n d
(2.112)-(2.114) as 195.4, 0.8557, a n d 1.481, respectively. F r o m
TABLE 3.26--Estimation of hydrogen content of petroleum fraction in Example 3.17.
Method Riazi Goossens ASTMD 3343 Jenkins-Walsh
%H, calc. 12.8 12.6 12.9 13.1
AD,% 0.1 0.3 0 0.2
Erroff
A A D % MAD%
0.01-1.6 0.09 0.7
0.07-6.2 0.24 1.6
0.01-6.2 0.15 1.6
Eqs. (2.15) a n d (3.50) we get Ri -- 1.0534, m - - 1.2195. Since
M < 250, Eq. (3.97) is used to e s t i m a t e the sulfur content as
%S = 1. i % versus the e x p e r i m e n t a l value of 0.8%. Therefore,
the e r r o r is calculated as follows: 1.1%-0.8% --- 0.3%.
To calculate %C a n d % H from e x p e r i m e n t a l data, values of
CH = 6.69 with %S = 0.8 are used in Eqs. (3.90) a n d (3.91).
This w o u l d result in %C -- 86.3 a n d %H -- 12.9. According to
the general m e t h o d p r e s e n t e d in this b o o k (author's p r o p o s e d
method), CH is calculated f r o m Eq. (2.120) as CH = 6.75 a n d
with e s t i m a t e d value of sulfur content as %S -- 1.1, %C a n d
%H are c a l c u l a t e d f r o m Eqs. (3.90) a n d (3.91) as %C = 86.1
a n d %H = 12.8. In use of Goossens m e t h o d t h r o u g h Eq. (3.92),
e s t i m a t e d values of n, d, a n d M are r e q u i r e d w h e r e M s h o u l d
be e s t i m a t e d f r o m Eq. (2.55) as M = 190. F o r this m e t h o d %C
m a y be calculated from Eq. (3.88) if %S is known. A s u m m a r y
o r results for calculation of % H with AD for various m e t h o d s
is given in Table 3.26. The Goossens m e t h o d gives the high-
est e r r o r b e c a u s e all i n p u t d a t a r e q u i r e d are p r e d i c t e d values.
The ASTM m e t h o d gives the s a m e value as e x p e r i m e n t a l value
b e c a u s e the e x p e r i m e n t a l values on all the i n p u t p a r a m e t e r s
r e q u i r e d in Eq. (3.93) are available in this p a r t i c u l a r example.
However, in m a n y cases a r o m a t i c c o n t e n t o r c o m p l e t e distilla-
tion curve as r e q u i r e d b y the ASTM m e t h o d are not available.
The general m e t h o d of a u t h o r p r e s e n t e d in this section b a s e d
on calculation of CH a n d %S gives g o o d results a l t h o u g h 50%
ASTM distillation t e m p e r a t u r e a n d specific gravity have b e e n
used as the only available data. t
Example 3.18--A p e t r o l e u m cut has the boiling r a n g e of 370-
565~ a n d is p r o d u c e d from a c r u d e oil from Danish N o r t h
Sea fields (Ref. [46], p. 353). The nitrogen content of c r u d e
is 1235 p p m . Calculate n i t r o g e n content of the fraction a n d
c o m p a r e with the e x p e r i m e n t a l value of 1625 p p m .
Solution--Tb = (370 + 565)/2 = 467.5~ = 740.6 K. T = Tb/
1000 = 0.7406. S u b s t i t u t i n g T in Eq. (3.99) gives %N2 in cut ----
1.23 x 1235 = 1525. The p e r c e n t relative deviation with the
e x p e r i m e n t a l value is - 6 % . This is relatively a g o o d predic-
tion, b u t n o r m a l l y larger errors are o b t a i n e d especially for
lighter cuts. t
3.6 P R E D I C T I O N OF O T H E R P R O P E R T I E S
In this section, predictive m e t h o d s for s o m e i m p o r t a n t p r o p -
erties that are useful to d e t e r m i n e the quality of certain
p e t r o l e u m p r o d u c t s a r e presented. S o m e of these proper-
ties such as flash p o i n t or p o u r p o i n t are useful for safety
Bureau of Mines Eq. (3.97)
13.1 12.7
0.2 0.2
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS
consideration or storage and transportation of products. One
of the most important properties of petroleum products re-
lated to volatility after the boiling point is vapor pressure.
For petroleum fractions, vapor pressure is measured by the
method of Reid. Methods of prediction of true vapor pressure
of petroleum fractions are discussed in Chapter 7. However,
Reid vapor pressure and other properties related to volatility
are discussed in this section. The specific characteristics of
petroleum products that are considered in this part are flash,
pour, cloud, freezing, aniline, and smoke points as well as car-
bon residue and octane number. Not all these properties apply
to every petroleum fraction or product. For example, octane
number applies to gasoline and engine type fuels, while car-
bon residue is a characteristic of heavy fractions, residues,
and crude oils. Freezing, cloud, and pour points are related
to the presence of heavy hydrocarbons and are characteristics
of heavy products. They are also important properties under
very cold conditions. Predictive methods for some of these
properties are rare and scatter. Some of these methods are
developed based on a limited data and should be used with
care and caution.
3.6.1 Properties Related to Volatility
Properties that are related to volatility of petroleum fraction
are boiling point range, density, Reid vapor pressure, and flash
point. Prediction of boiling point and density of petroleum
fractions have been discussed earlier in this chapter. In this
FIG. 3.27--Apparatus to measure RVP of
petroleum products by ASTM D 323 test method
(courtesy of KISR).
131
part, methods of prediction of vapor pressure, fuel vapor liq-
uid (V/L) ratio, fuel volatility index, and flash points are pre-
sented.
3.6.1. i Reid Vapor Pressure
Reid vapor pressure is the absolute pressure exerted by a mix-
ture at 37.8~ (311 K or 100~ at a vapor-to-liquid volume
ratio of 4 [4]. The RVP is one of the important properties of
gasolines and jet fuels and it is used as a criterion for blend-
ing of products. RVP is also a useful parameter for estimation
of losses from storage tanks during filling or draining. For
example, according to Nelson method losses can be approxi-
mately calculated as follows: losses in vol% = (14.5 RVP - 1)/6,
where RVP is in bar [24, 66]. The apparatus and procedures
for standard measurement of RVP are specified in ASTM
D 323 or IP 402 test methods (see Fig. 3.27). In general, true
vapor pressure is higher than RVP because of light gases dis-
solved in liquid fuel. Prediction of true vapor pressure of pure
hydrocarbons and mixtures is discussed in detail in Chapter 7
(Section 7.3). The RVP and boiling range of gasoline governs
ease of starting, engine warm-up, mileage economy, and ten-
dency toward vapor lock [63]. Vapor lock tendency is directly
related to RVP and at ambient temperature of 21~ (70~
the maximum allowable RVP is 75.8 kPa (11 psia), while this
limit at 32~ (90~ reduces to 55.2 kPa (8 psia) [63]. RVP
can also be used to estimate true vapor pressure of petroleum
fractions at various temperatures as shown in Section 7.3.
True vapor pressure is important in the calculations related
to losses and rate of evaporation of liquid petroleum prod-
ucts. Because RVP does not represent true vapor pressure,
the current tendency is to substitute RVP with more modern
and meaningful techniques [24]. The more sophisticated in-
struments for measurement of TVP at various temperatures
are discussed in ASTM D 4953 test method. This method
can be used to measure RVP of gasolines with oxygenates
and measured values are closer to actual vapor pressures
E4, 24].
As will be discussed in Chapters 6 and 7, accurate calcu-
lation of true vapor pressure requires rigorous vapor liquid
equilibrium (VLE) calculations through equations of state.
The API-TDB [2] method for calculation of RVP requires a
tedious procedure with a series of flash calculations through
Soave cubic equation of state. Simple relations for estima-
tion of RVP have been proposed by Jenkins and White and
are given in Ref. [61]. These relations are in terms of tem-
peratures along ASTM D 86 distillation curve. An example of
these relations in terms of temperatures at 5, i0, 30, and 50
vol% distilled is given below:
RVP = 3.3922 - 0.02537(T5) - 0.070739(T10) + 0.00917(T30)
- 0.0393(Ts0) + 6.8257 x 10-4(T10) 2
(3.100)
where all temperatures are in~ and RVP is in bar. The diffi-
culty with this equation is that it requires distillation data up
to 50% point and frequently large errors with negative RVP
values for heavier fuels have been observed. Another method
for prediction of RVP was proposed by Bird and Kimball [61 ].
In this method the gasoline is divided into a number (i.e.,
28) of cuts characterized by their average boiling points. A
132 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
blending RVP of each cut is then calculated by the following
equation:
7.641
Bi=
exp (0.03402Tbi + 0.6048)
i=28
(3.101) Pa = ~ BiXvi
i=l
f -- 1.0 + 0.003744 (VAPB - 93.3)
RVP = fPa
where B~ = RVP is the blending number for cut i and T b i =
normal boiling point of cut i in ~ Xvi is the volume fraction
of cut i, VABP is the volume average boiling point in~ and
RVP is the Reid vapor pressure in bars. The constants were ob-
tained from the original constants given in the English units.
The average error for this method for 51 samples was 0.12
bar orl.8 psi [67, 68].
Recently some data on RVP of gasoline samples have been
reported by Hatzioznnidis et al. [69]. They measured vapor
pressure according to ASTM D 5191 method and related to
RVP. They also related their measured vapor pressure data to
TVP thus one can obtain RVP from TVP, but their relations
have not been evaluated against a wide range of petroleum
fractions. Other relations for calculation of TVP from RVP
for petroleum fractions and crude oils are given in Section
7.3.3. TVP at 100~ (311 K) can be estimated from Eq. (3.33)
as
(3.102) logl0(TVP)100=3.204x 1-4x l~_-~r b
where Tb is the normal boiling point in K and TVP100 is the
true vapor pressure at 100~ (311 K). Once TVP is calculated
it may be used instead of RVP in the case of lack of sufficient
data. When this equation is used to estimate RVP of more than
50 petroleum products an average error of 0.13 bar (~ 1.9 psi)
and a maximum error of 5.9 psi were obtained [67, 68].
RVP data on 52 different petroleum products (light and
heavy naphthas, gasolines, and kerosenes) from the Oil and
Gas Journal data bank [46] have been used to develop a sim-
ple relation for prediction of RVP in terms of boiling point
and specific gravity in the following form [67]:
RVP = Pc exp(Y)
r=- X (TbSG'~
',--g--~ / (1 - Tr)s
X = -276.7445 + 0.06444Tb + 10.0245SG - 0.129TbSG
9968.8675
+ + 44.6778 In Tb + 63.6683 In SG
TbSG
T~ = 311/Tc
(3.103)
where Tb is the mid boiling point and Tc is the pseudocritical
temperature of the fraction in kelvin. Pc is the pseudocritical
pressure and RVP is the Reid vapor pressure in bars. The basis
for development of this equation was to use Miller equation
for TVP and its application at 311 K (100~ The Miller equa-
tion (Eq. 7.13) is presented in Section 7.3.1. The constants
of vapor pressure correlation were related to boiling point
and specific gravity of the fraction. Critical temperature and
pressure may be estimated from Tb and SG using methods
presented in Chapter 2. This equation is based on data with
RVP in the range of 0.0007-1.207 bar (0.01-17.5 psia), normal
boiling point range of 305-494 K, and specific gravity range
of 0.65-1.08. The average absolute deviation for 52 samples is
0.061 bar (0.88 psia). The above equation may be used for cal-
culation of RVP to determine quality characteristics of a fuel.
The calculated RVP value should not be used for calculation
of TVP when very accurate values are needed. (Appropriate
methods for direct estimation of TVP of petroleum fractions
are discussed in Section 7.3.3.) Vapor pressure of a petroleum
mixture depends on the type of its constituents and with use
of only two bulk properties to predict RVP is a difficult task.
This equation is recommended for a quick and convenient
estimation of RVP, but occasionally large errors may be ob-
tained in use of this equation. For more accurate estimation
of RVP the sophisticated method suggested in the API-TDB
[2] may be used. In this method RVP is calculated through a
series of vapor-liquid-equilibrium calculations.
RVP is one of the main characteristics that is usually used
to blend a fuel with desired specifications. The desired RVP
of a gasoline is obtained by blending naphtha with n-butane
(M = 58, RVP = 3.58 bar or 52 psia) or another pure hydro-
carbon with higher RVPs than the original fuel. For condi-
tions where RVP should be lowered (hot weather), heavier
hydrocarbons with lower RVP are used for blending purposes.
RVP of several pure hydrocarbons are given as follows: i-C4:
4.896 bar (71 psia); n-C4:3.585 (52); i-Cs:1.338 (19.4); n-Cs:
1.0135 (14.7); i-C6:0.441 (6.4); n-C6:0.34 (5.0); benzene: 0.207
(3.0); and toluene: 0.03 (0.5), where all the numbers inside the
parentheses are in psia as given in Ref. [63]. However in the
same reference in various chapters different values of RVP
for a same compound have been used. For example, values of
4.14 bar (60 psi) for n-C4, 1.1 bar (16 psi) for n-Cs, and 0.48
bar (7 psi) for i-C6 are also reported by Gary and Handwerk
[63]. They also suggested two methods for calculation of RVP
of a blend when several components with different RVPs are
blended. The first method is based on the simple Kay's mixing
rule using mole fraction (Xr~) of each component [63]:
(3.104) RVP(blend) = E Xmi(RVP)/
i
where (RVP)i is the RVP of component i in bar or psia. The
second approach is to use blending index for RVP as [63]:
(RVPBI)i = (RVP)] 2s
(3.105) RVPBI (blend) = EXvi(RVPBI)i
i
RVP (blend) = [RVPBI (blend)] ~
where (RVPBI)/ is the blending index for (RVP)i and Xvi is
the volume fraction of component i. Both units of bar or
psia may be used in the above equation. This relation was
originally developed by Chevron and is also recommended
in other industrial manuals under Chevron blending number
[61]. Equations (3.104) and (3.105) may also be applied to
TVP; however, methods of calculation of TVP of mixtures are
discussed in Section 7.3 through thermodynamic relations.
Example 3.19--Estimate RVP of a gasoline sample has
molecular weight of 86 and API gravity of 86.
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S
Solution--API = 86 and M = 86. From Eq. (2.4), SG = 0.65
and from Eq. (2.56), Tb = 338 K. Since only Tb and SG are
known, Eq. (3.103) is used to calculate the RVP. From Eqs.
(2.55) and (2.56) we get Tc -- 501.2 K a n d Pc = 28.82 bar. From
Eq. (3.103), Tr = 0.6205, X = 1.3364, and Y = -3.7235. Thus
we calculate RVP = 0.696 bar or 10.1 psia. The experimental
value is I 1.1 psia [63]. # Flash point of petroleum fractions is the lowest temperature
3.6.1.2 V/L Ratio and Volatility Index
Once RVP is known it can be used to determine two other
volatility characteristics, namely vapor liquid ratio (V/L) and
fuel volatility index (FVI), which are specific characteristics
of spark-ignition engine fuels such as gasolines. V/L ratio is a
volatility criterion that is mainly used in the United States and
Japan, while FVI is used in France and Europe [24]. The V/L
ratio at a given temperature represents the volume of vapor
formed per unit volume of liquid initially at 0~ The proce-
dure of measuring V/L ratio is standardized as ASTM D 2533.
The volatility of a fuel is expressed as the temperature levels at
which V/L ratio is equal to certain values. Usually V/L values
of 12, 20, and 36 are of interest. The corresponding tempera-
tures may be calculated from the following relations [24]:
T(V/L)12 m_ 8 8 . 5 -- 0 . 1 9 E 7 0 - 42.5 RVP
(3.106) T(v/L)2o= 90.6 - 0.25E70 - 39.2 RVP
T~v/L)36 = 94.7 -- 0.36E70 - 32.3 RVP
where T(v/L)x is the temperature in ~ at which V/L = x.
Parameter E70 is the percentage of volume distilled at 70~
E70 and RVP are expressed in percent distilled and bar,
respectively. Through Lagrange interpolation formula it is
possible to derive a general relation to determine temper-
ature for any V/L ratio. E70 can be calculated through a
distribution function for distillation curve such as Eq. (3.35)
in which by rearrangement of this equation we get
(3.107) E70 = 1 0 0 - 100 exp -
E \
where To is the initial boiling point in kelvin and together
To
] (15.19 psia).
with parameters A and B can be determined from the
method discussed in Section 3.2.3. Another simple relation
to calculate T(V/L)20 is given in terms of RVP and distillation
temperatures at 10 and 50% [61]:
(3.108) T(V/L)2O= 52.5 + 0.2T10 + 0.17T50 - 33 RVP
where T10 and Ts0 are temperatures at 10 and 50 vol% distilled
on the ASTM D 86 distillation curve. All temperatures are
in~ and RVP is in bar. For cases that T10 is not available it
may be estimated through reversed form of Eq. (3.17) with T50
and SG. Several petroleum refining companies in the United
States such as Exxon and Mobil use the critical vapor locking
index (CVLI), which is also related to the volatility index [61 ].
(3.109) CVLI -- 4.27 + 0.24E70 + 0.069 RVP
The fuel volatility index is expressed by the following relation
[24]:
(3.110) FVI = 1000 RVP + 7E70
where RVP is in bar. FV/is a characteristic of a fuel for its per-
formance during hot operation of the engine. In France, spec-
133
ifications require that its value be limited to 900 in summer,
1000 in fall/spring, and 1150 in the winter season. Automobile
manufacturers in France require their own specifications
that the value of FVI not be exceeded by 850 in summer [24].
3.6.1.3 Flash Point
at which vapors arising from the oil will ignite, i.e. flash,
when exposed to a flame under specified conditions. There-
fore, the flash point of a fuel indicates the maximum tem-
perature that it can be stored without serious fire hazard.
Flash point is related to volatility of a fuel and presence of
light and volatile components, the higher vapor pressure cor-
responds to lower flash points. Generally for crude oils with
RVP greater than 0.2 bar the flash point is less than 20~ [24].
Flash point is an important characteristics of light petroleum
fractions and products under high temperature environment
and is directly related to the safe storage and handling of
such petroleum products. There are several methods of de-
termining flash points of petroleum fractions. The Closed
Tag method (ASTM D 56) is used for petroleum stocks with
flash points below 80 ~C (175 ~ The Pensky-Martens method
(ASTM D 93) is used for all petroleum products except waxes,
solvents, and asphalts. Equipment to measure flash point ac-
cording to ASTM D 93 test method is shown in Fig. 3.28.
The Cleveland Open Cup method (ASTM D 92) is used for
petroleum fractions with flash points above 80~ (175~ ex-
cluding fuel oil. This method usually gives flash points 3-6~
higher than the above two methods [61]. There are a number
of correlations to estimate flash point of hydrocarbons and
petroleum fractions.
Buffer et al. [70] noticed that there is a linear relationship
between flash point and normal boiling point of hydrocar-
bons. They also found that at the flash point temperatures, the
product of molecular weight (M) and vapor pressure (pvap) for
pure hydrocarbons is almost constant and equal to 1.096 bar
(3.111) MP vap = 1.096
Another simple relation for estimation of flash point of hydro-
carbon mixtures from vapor pressure was proposed by Walsh
and Mortimer [71].
(3.112) TF = 231.2 -- 40 logP vav
where pvap is the vapor pressure at 37.8~ (100~ in bar and
TF is the flash point in kelvin. For simplicity RVP may be used
for pwp. Methods of calculation of vapor pressure are dis-
cussed in Chapter 7. Various oil companies have developed
special relations for estimation of flash points of petroleum
fractions. Lenoir [72] extended Eq. (3.100) to defined mix-
tures through use of equilibrium ratios.
The most widely used relation for estimation of flash point
is the API method [2], which was developed by Riazi and
Daubert [73]. They used vapor pressure relation from Clasius-
Clapeyron (Chapter 6) together with the molecular weight
relation form Eq. (2.50) in Eq. (3.111) to develop the following
relation between flash point and boiling point:
(3.113) 1/TF=a+b/Tb+clnTb+dlnSG
where Tb is the normal boiling point of pure hydrocarbons.
It was observed that the coefficient d is very small and TF is
134 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

FIG, 3.28---Equipment for measurement of flash point of petroleum fractions by ASTM

D 93 test method (courtesy of Chemical Engineering Department at Kuwait University),

nearly independent of specific gravity. Based on data from (3.114) is adopted by the API as the standard method to esti-
pure hydrocarbons and some petroleum fractions, the con- mate flash point of petroleum fractions [2]. It was shown that
stants in Eq. (3.113) were determined as Eq. (3.114) can be simplified into the following linear form
1 2.84947 [73]:
(3.114) ~FF = --0.024209+ T1----~+3.4254 • 10 -3 lnTm
(3.115) TF = 15.48 + 0.70704/'1o
where for pure hydrocarbons T10 is normal boiling point,
while for petroleum fractions it is distillation temperature at where both Tx0 and TF are in kelvin. This equation is applica-
10 vol% vaporized (ASTM D 86 at 10%) and it is in kelvin. TF is ble to fractions with normal boring points (i.e., ASTM D 86
the flash point in kelvin determined from the ASTM D 93 test temperature at 50%) less than 260~ (500~ For such light
method (Pensky-Martens closed cup tester). This equation is fractions, Eq. (3.115) is slightly more accurate than Eq.
presented in Fig. 3.29 for a quick and convenient estimate (3.t14). For heavier fractions Eq. (3.114) should be used.
of flash point. For 18 pure hydrocarbons and 39 fractions, There are some relations in the literature that correlate flash
Eq. (3.114) predicts flash points with an average absolute de- points to either the initial boiling point (T10) or the distillation
viation (AD) of 6.8~ (12~ while Eq. (3.111) predicts the temperature at 50% point (T50). Such correlations are not ac-
flash points with AD of 18.3~ curate over a wide range of fractions, especially when they are
Equation (3.114) should be applied to fractions with nor- applied to fractions not used in obtaining their coefficients.
mal boring points from 65 to 590~ 150-1100~ Equation Generally reported initial boiling points for petroleum frac-
tions are not reliable and if mid boiling point temperature is
used as the characteristics boiling point it does not truly rep-
150 resent the boiling point of light components that are initially
being vaporized. For this reason the correlations in terms of
o ~176 ILILL distillation temperature at 10% point (7"10)are more accurate
than the other correlations for estimation of flash points of
petroleum fractions. Flash points of petroleum fractions may
also be estimated from the pseudocomponent method using
the PNA composition and values of flash points of pure hy-
drocarbons from Table 2.2. However, volumetric averaging
of component flash point through Eq. (3.40) generally over-
-50 predicts the flash point of the blend and the blending index
0 100 200 300 400 approach described below should be used to estimate flash
point of defined mixtures.
ASTM 10% Temperature,~
If the flash point of a petroleum fraction or a petroleum
FIG. 3.29--Prediction of flash product does not meet the required specification, it can be
point of petroleum fractions from adjusted by blending the fraction with other compounds hav-
Eq. (3.114). ing different flash points. For example in hot regions where
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS
the temperature is high, heavy hydrocarbons may be added
to a fraction to increase its flash point. The flash point of the
blend should be determined from the flash point indexes of
the components as given below [74]:
2414
(3.116) lOgl0 BIF = -6.1188 + - -
TF - 42.6
where log is the logarithm of base 10, BIF is the flash point
blending index, and TF is the flash point in kelvin. Once BIF
is determined for all components of a blend, the blend flash
point index (BIB) is determined from the following relation:
(3.117) BIB = ~_x,,iBIi
where xvi is the volume fraction and BIi is the flash point
blending index of component i. As it will be shown later, the
blending formula by Eq. (3.117) will be used for several other
properties. Once Bits is calculated it should be used in Eq.
(3.116) to calculate the flash point of the blend, TFB. Another
relation for the blending index is given by Hu-Burns [75]:
(3.118) Tl/X
B I F = ~F
where TF is the flash point in kelvin and the best value of x
is -0.06. However, they suggest that the exponent x be cus-
tomized for each refinery to give the best results [61]. The
following example shows application of these methods.
Example 3.20---A kerosene product with boiling range of
175-260~ from Mexican crude oil has the API gravity of 43.6
(Ref. [46], p. 304). (a) Estimate its flash point and compare
with the experimental value of 59~ (b) For safety reasons it
is required to have a m i n i m u m flash point of 65~ to be able
to store it in a hot summer. How much n-tetradecane should
be added to this kerosene for a safe storage?
Solutions(a) To estimate flash point we use either Eq.
(3.114) or its simplified form Eq. (3.15), which require ASTM
10% temperature, T10. This temperature may be estimated
from Eq. (3.17) with use of specific gravity, SG = 0.8081, and
ASTM 50% temperature, Ts0. Since complete ASTM curve is
not available it is assumed that the mid boiling point is the
same as Ts0; therefore, T50 = 217.5~ and from Eq. (3.17) with
coefficients in Table 3.4,/'10 = 449.9 K. Since Ts0 is less than
260~ Eq. (3.115) can be used for simplicity. The result is
TF = 60.4~ which is in good agreement with the experimen-
tal value of 59~ considering the fact that an estimated value
of ASTM 10% temperature was used.
(b) To increase the flash point from 59 to 65~ n-Ct4 with
flash point of 1000C (Table 2.2) is used. If the volume frac-
tion of ~/-C14 needed is shown by Xadd, then using Eq. (3.117)
we have BIFB = (1 --Xadd) X BIvK + Xadd X BIVadd where B I ~ ,
BIFK, and BIradd are the blending indexes for flash points
of final blend, kerosene sample, and the additive (n-C14),
respectively. The blending indexes can be estimated from
Eq. (3.116) as 111.9, 165.3, and 15.3, respectively, which re-
sult in xaaa = 0.356. This means that 35.6% in volume of rt-C14
is required to increase the flash point to 65~ If the blending
indexes are calculated from Eq. (3.118), the amount of r/-C14
required is 30.1%. , in the range of 0-10~ (0-18~ as shown in Table 3.27. The
135
3.6.2 P o u r P o i n t
The pour point of a petroleum fraction is the lowest tempera-
ture at which the oil will pour or flow when it is cooled without
stirring under standard cooling conditions. Pour point repre-
sents the lowest temperature at which an oil can be stored
and still capable of flowing under gravity. Pour point is one
of low temperature characteristics of heavy fractions. When
temperature is less than pour point of a petroleum product it
cannot be stored or transferred through a pipeline. Test pro-
cedures for measuring pour points of petroleum fractions are
given under ASTM D 97 (ISO 3016 or IP 15) and ASTM D 5985
methods. For commercial formulation of engine oils the pour
point can be lowered to the limit of - 2 5 and -40~ This is
achieved by using pour point depressant additives that inhibit
the growth of wax crystals in the oil [5]. Presence of wax and
heavy compounds increase the pour point of petroleum frac-
tions. Heavier and more viscous oils have higher pour points
and on this basis Riazi and Daubert [73] used a modified ver-
sion of generalized correlation developed in Chapter 2 (Eq.
2.39) to estimate the pour point of petroleum fractions from
viscosity, molecular weight, and specific gravity in the follow-
ing form:
Tp = 130.47[SG 297~ x [M (~176
F (0.310331-0.32834SG)1
(3.119) x lP38(100) J
where Tp is the pour point (ASTM D 97) in kelvin, M is the
molecular weight, and v38o00) is the kinematic viscosity at
37.8~ (100~ in eSt. This equation was developed with data
on pour points of more than 300 petroleum fractions with
molecular weights ranging from 140 to 800 and API gravities
from 13 to 50 with the AAD of 3.9~ [73]. This method is also
accepted by the API and it is included in the API-TDB since
1988 [2] as the standard method to estimate pour point of
petroleum fractions. As suggested by Hu and Burns [75, 76],
Eqs. (3.117) and (3.118) used for blending index of flash point
can also be used for pour point blending index (TpB) with
x = 0.08 :
(3.120) BIp = T1/0.08
~p
where Tp is the pour point of fraction or blend in kelvin. The
AAD of 2.8~ is reported for use of Eqs. (3.117) and (3.120)
to estimate pour points of 47 blends [76].
3.6.3 C l o u d P o i n t
The cloud point is the lowest temperature at which wax crys-
tals begin to form by a gradual cooling under standard con-
ditions. At this temperature the oil becomes cloudy and the
first particles of wax crystals are observed. The standard pro-
cedure to measure the cloud point is described under ASTM
D 2500, IP 219, and ISO 3015 test methods. Cloud point
is another cold characteristic of petroleum oils under low-
temperature conditions and increases as molecular weight of
oil increases. Cloud points are measured for oils that con-
tain paraffins in the form of wax and therefore for light frac-
tions, such as naphtha or gasoline, no cloud point data are
reported. Cloud points usually occur at 4-5~ (7 to 9~ above
the pour point although the temperature differential could be
136 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

TABLE 3.27--Cloud and pour points and their differences for some petroleum products.
Fraction API gravity T~,,~C TCL,~C Tp -TcL, ~
Indonesian Dist. 33.0 -43.3 -53.9 10.6
Australian GO 24.7 -26.0 -30.0 4.0
Australian HGO 22.0 -8.0 -9.0 1.0
Abu Dhabi LGO 37.6 -19.0 -27.0 8.0
Abu Dhabi HGO 30.3 7.0 2.0 5.0
Abu Dhabi Disst. 21.4 28.0 26.0 2.0
Abu Dhabi Diesel 37.4 - 12.0 - 12.0 0.0
Kuwaiti Kerosene 44.5 -45.0 -45.0 0.0
Iranian Kerosene 44.3 -46.7 -46.7 0.0
Iranian Kerosene 42.5 -40.6 -48.3 7.8
Iranian GO 33.0 - 11.7 - 14.4 2.8
North Sea GO 35.0 6.0 6.0 0.0
Nigerian GO 27.7 -32.0 -33.0 1.0
Middle East Kerosene 47.2 -63.3 -65.0 1.7
Middle East Kerosene 45.3 -54.4 -56.7 2.2
Middle East Kerosene 39.7 -31.1 -34.4 3.3
Middle East Disst. 38.9 -17.8 -20.6 2.8
Source: Ref. [46].
Tp: pour point; TcL:cloud point.

difference between cloud and p o u r point depends on the na-
ture of oil and there is no simplified correlation to predict this
difference. Cloud point is one of the important characteris-
tics of crude oils under low-temperature conditions. As tem-
perature decreases below the cloud point, formation of wax
crystals is accelerated. Therefore, low cloud point products
are desirable u n d e r low-temperature conditions. Wax crys-
tals can plug the fuel system lines and filters, which could
lead to stalling aircraft and diesel engines u n d e r cold condi-
tions. Since cloud point is higher than p o u r point, it can be
considered that the knowledge of cloud point is more impor-
tant than the p o u r point in establishing distillate fuel oil spec-
ifications for cold weather usage [61]. Table 3.27 shows the
difference between cloud and p o u r points for some petroleum
products. Cloud and pour points are also useful for predict-
ing the temperature at which the observed viscosity of an oil
deviates from the true (newtonian) viscosity in the low tem-
perature range [7]. The a m o u n t of n-paraffins in petroleum
oil has direct effect on the cloud point of a fraction [8]. Pres-
ence of gases dissolved in oil reduces the cloud point which is
desirable. The exact calculation of cloud point requires solid-
liquid equilibrium calculations, which is discussed in Chap-
ter 9. The blending index for cloud point is calculated from
the same relation as for p o u r point through Eq. (3.118) with
x = 0.05 :
(3.121) BIcL = T~/~176
where TCL is the cloud point of fraction or blend in kelvin.
Accuracy of this m e t h o d of calculating cloud point of blends is
the same as for the pour point (AAD of 2.8~ Once the cloud
point index for each c o m p o n e n t of blend, BIcLi, is determined
through Eq. (3.12 t), the cloud point index of the blend, Blczs,
is calculated through Eq. (3.117). Then Eq. (3.121) is used in
its reverse form to calculate cloud point of the blend from its
cloud point index [76].
3.6.4 Freezing Point
Freezing point is defined in Section 2.1.9 and freezing points
of pure hydrocarbons are given in Table 2.2. For a petroleum
fraction, freezing point test involves cooling the sample until a
slurry of crystals form throughout the sample or it is the tem-
perature at which all wax crystals disappear on rewarming
the oil [61]. Freezing point is one of the important charac-
teristics of aviation fuels where it is determined by the pro-
cedures described in ASTM D 2386 (U.S.), IP 16 (England),
and NF M 07-048 (France) test methods. M a x i m u m freezing
point of jet fuels is an international specification which is re-
quired to be at - 4 7 ~ (-53~ as specified in the "Aviation
Fuel Quantity Requirements for Jointly Operated Systems"
[24]. This m a x i m u m freezing point indicates the lowest tem-
perature that the fuel can be used without risk of separation
of solidified hydrocarbons (wax). Such separation can result
in the blockage in fuel tank, pipelines, nozzles, and filters [61 ].
Walsh-Mortimer suggest a t h e r m o d y n a m i c model based on
the solubility of n-paraffin hydrocarbons in a petroleum mix-
ture to determine the freezing point [71]. Accurate determi-
nation of freezing point requires accurate knowledge of the
composition of a fuel which is normally not known. However,
the m e t h o d of determination of carbon n u m b e r distribution
along with solid-liquid equilibrium can be used to determine
freezing points of petroleum fractions and crude oils as will be
discussed in Chapter 9. A simpler but less accurate method to
determine freezing points of petroleum fractions is t h r o u g h
the p s e u d o c o m p o n e n t approach as shown in the following
example.
Example 3.21--A kerosene sample produced from a crude
oil from North Sea Ekofisk field has the boiling range of 150-
204.4~ (302--400~ and API gravity of 48.7. Estimate the
freezing point of this kerosene and c o m p a r e with the experi-
mental value of - 6 5 ~ (-85~
Solution--The mid boiling point is Tb = 177.2~ and the spe-
cific gravity is SG = 0.785. We use the method of pseudo-
c o m p o n e n t using predicted composition. F r o m Eq. (2.50),
M = 143 and since M > 143, we use Eqs. (3.77), (3.78), and
(3.72) to predict Xp, XN, and XA, respectively. F r o m Eqs. (2.114)
and (2.115), n = 1.439 and from Eq. (3.50), m = -5.1515. Us-
ing SG and m, we calculate the PNA composition as xp =
0.457,xN = 0.27, and XA = 0.273. F r o m Eqs. (3.41)-(3-43),
Mp = 144.3, MN = 132.9, and MA = 129.3. Using Eq. (2.42)
for prediction of the freezing point for different families we
get TFp = 242.3, TFN = 187.8, and TrA = 178.6 K. Using Eq.
(3.40) we get TF = 210.2 K or -63.1~ versus the measured
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS
FIG. 3.30~Apparatus to measure aniline point of petroleum fuels by
ASTM D 611 test method (courtesy of KISR).
value of -65~ The result is quite satisfactory considering
minimum data on Tb and SG are used as the only available
parameters.
3.6.5 Aniline P o i n t
Aniline point of a petroleum fraction is defined as the mini-
m u m temperature at which equal volumes of aniline and the
oil are completely miscible. Method of determining aniline
point of petroleum products is described under ASTM D 611
test method and the apparatus is shown in Fig. 3.30. Aniline
point indicates the degree of aromaticity of the fraction. The
higher the aniline point the lower aromatic content. For this
reason aromatic content of kerosene and jet fuel samples may
be calculated from aniline point [59]:
(3.122) %A = 692.4 + 12.15(SG)(AP) - 794(SG) - 10.4(AP)
where %A is the percent aromatic content, SG is the specific
gravity, and AP is the aniline point in~ There are a number
of methods to estimate aniline point of petroleum fractions.
We discuss four methods in this section,
3.6.5.1 Winn Method
Aniline point can be estimated from Winn nomograph
(Fig. 2.14) using Tb and SG or M and SG as the input pa-
rameters.
3.6.5.2 Walsh-Mortimer
The aniline point can be calculated from the following rela-
tion [61,71]:
(3.123) AP = -204.9 - 1.498Cs0 +
100.5c /03
SG
where AP is the aniline point in ~ and C5o is the carbon
number of n-paraffin whose boiling point is the same as the
mid boiling point of the fraction. C50 may be calculated from
the following relation:
Mp-14
(3.124) C50 - - -
2 vors auto-ignition. The ignition delay period can be evaluated
137
in which Me is the molecular weight of n-paraffin whose boil-
ing point is the same as mid boiling point of the fraction which
can be determined from Eq. (3.41).
3.6.5.3 Linden Method
This relation is a mathematical representation of an earlier
graphical method and is given as [73]
(3.125) AP = -183.3 + 0.27(API)T~/3 + 0.317Tb
where AP is in ~ Tb is the mid boiling point in kelvin and
API is API gravity. The blending index for aniline point may be
calculated from the following relation developed by Chevron
Research [61 ]:
(3.126) BIAp = 1.124[exp (0.00657AP)]
where AP is in ~ and BIApis the blending index for the aniline
point. Once the blending indexes of components of a blend are
determined, Eq. (3.117) should be used to calculate blending
index for aniline point of the blend.
3.6.5.4 Albahri et al. Method
Most recently Mbahri et al. [68] developed predictive meth-
ods for determination of quality of petroleum fuels. Based on
the idea that aniline point is mainly related to the aromatic
content of a fuel, the following relation was proposed:
(3.127) AP = -9805.269(Ri) + 711.85761(SG) + 9778.7069
where AP is in ~ and Ri is defined by Eq. (2.14). Equations
(3.123), (3.125), and (3.127) were evaluated against data on
aniline points of 300 fuels with aniline point range: 45-107~
boiling range: 115-545~ and API gravity range of 14-56. The
average absolute deviation (AAD) for Eq. (3.127) was 2.5~
while for Eqs. (3.123) and (3.125) the errors were 4.6 and
6.5~ respectively [68]. Error distribution for Eq. (3.127) is
shown in Fig. 3.31.
3.6.6 Cetane N u m b e r a n d Diesel I n d e x
For diesel engines, the fuel must have a characteristic that fa-
138 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
120.0
110.0
100.0
90.0
fl 80.0 . 9
70.0 , ,
60.0 . -~. ~"
"%Z "
~o.o "/:
40.0 /
30.0
30.0 40.0 50.0 60.0 70.0 80.0
Experimental Aniline Point, ~
FIG. 3.31--Error distribution for prediction of aniline point
from Eq. (3.127). Taken with permission from Ref. [68].
by the fuel characterization factor called cetane number (CN).
The behavior of a diesel fuel is measured by comparing
its performance with two pure hydrocarbons: n-cetane or
n-hexadecane (n-C16H34) which is given the number 100 and
a-methylnaphthalene which is given the cetane number of
0. A diesel fuel has a cetane number of 60 if it behaves
like a binary mixture of 60 vol% cetane and 40 vol% ~-
methylnaphthalene. In practice heptamethylnonane (HMN)
a branched isomer of n-cetane with cetane number of 15 is
used instead of a-methylnaphthalene [24, 61]. Therefore, in
practice the cetane number is defined as:
(3.128) CN = vol%(n-cetane) + 0.15(vo1% HMN)
The cetane number of a diesel fuel can be measured by the
ASTM D 613 test method. The shorter the ignition delay pe-
riod the higher CN value. Higher cetane number fuels re-
duce combustion noise and permit improved control of com-
bustion resulting in increased engine efficiency and power
output. Higher cetane number fuels tend to result in easier
starting and faster warm-up in cold weather. Cetane num-
ber requirement of fuels vary with their uses. For high speed
city buses in which kerosene is used as fuel the required CN
is 50. For premium diesel fuel for use in high speed buses
and light marine engines the required number is 47 while for
marine distillate diesel for low speed buses and heavy marine
engines the required cetane number is 38 [61]. In France the
minimum required CN of fuels by automotive manufacturers
is 50. The product distributed in France and Europe have CN
in the range of 48-55. In most Scandinavian countries, the
United States and Canada the cetane number of diesel fuels
are most often less than 50. Higher cetane number fuels in
addition to better starting condition can cause reduction in
air pollution [24].
Since determination of cetane number is difficult and
costly, ASTM D 976 (IP 218) proposed a method of calcu-
lation. Calculated number is called calculated cetane index
(CCI) and can be determined from the following relation:
CCI = 454.74 - 1641.416SG + 774.74SG 2
(3.129) - 0.554Ts0 + 97.083(log10 T50)2
where/'so is the ASTM D 86 temperature at 50% point in ~
Another characteristic of diesel fuels is called diesel index (DI)
defined as:
9
(3.130) OI =
(API)(1.8AF + 32)
100
which is a function of API gravity and aniline point in ~
Cetane index is empirically correlated to DI and AP in the
following form [24]:
(3.131) CI = 0.72DI + 10
(3.132) CI = AP - 15.5
where AP is in ~ Calculated cetane index (CI) is also related
to n-paraffin content (%NP) of diesel fuels in the following
90.0 100.0 110.0 120.0 from [87].
(3.133) %NP = 1.45CI - 57.5
The relation for calculation of cetane number blending in-
dex is more complicated than those for pour and cloud point.
Blending indexes for cetane number are tabulated in various
sources [61, 75]. Cetane number of diesel fuels can be im-
proved by adding additives such as 2-ethyl-hexyl nitrate or
other types of alkyl nitrates. Cetane number is usually im-
proved by 3-5 points once 300-1000 ppm by weight of such
additives is added [24]. Equation (3.129) suggested for cal-
culating cetane number does not consider presence of addi-
tives and for this reason calculated cetane index for some
fuels differ with measured cetane index. Generally, CCI is less
than measured CN and for this reason in France automobile
manufacturers have established minimum CN for both the
calculated CI (49) and the measured CN (50) for the quality
requirement of the fuels [24].
3.6.7 O c t a n e N u m b e r
Octane number is an important characteristic of spark en-
gine fuels such as gasoline and jet fuel or fractions that
are used to produce these fuels (i.e., naphthas) and it rep-
resents antiknock characteristic of a fuel. Isooctane (2,2,4-
trimethylpentane) has octane number of 100 and n-heptane
has octane number of 0 on both scales of RON and MON.
Octane number of their mixtures is determined by the vol%
of isooctane used. As discussed in Section 2.1.13, isoparaffins
and aromatics have high octane numbers while n-paraffins
and olefins have low octane numbers. Therefore, octane num-
ber of a gasoline depends on its molecular type composition
especially the amount of isoparaffins. There are two types of
octane number: research octane number (RON) is measured
under city conditions while motor octane number (MON)
is measured under road conditions. The arithmetic average
value of RON and MON is known as posted octane number
(PON). RON is generally greater than MON by 6-12 points,
although at low octane numbers MON might be greater than
RON by a few points. The difference between RON and MON
is known as sensitivity of fuel. RON of fuels is determined
 3. C H A R A C T E R I Z A T I O N OF PETROLEUM FRACTIONS 139

by ASTM D 908 and MON is measured by ASTM D 357 test 3.28--Octane numbers of some alcohols and
TABLE
methods. Generally there are three kinds of gasolines: regu- ethers (oxygenates).
Compound RON MON
lar, intermediate, and premium with PON of 87, 90, and 93,
Methanol 125-135 100-105
respectively. In France the minimum required RON for su- MTBE 113-I17 95-101
perplus gasoline is 98 [24]. Required RON of gasolines vary Ethanol 120-130 98-103
with parameters such as air temperature, altitude, humidity, ETBE 118-122 100-102
engine speed, and coolant temperature. Generally for every TBA 105-110 95-100
300 m altitude RON required decreases by 3 points and for TAME 110-114 96-100
M T B E : m e t h y l - t e r t i a r y - b u t y l ether; E T B E : e t h y l - t e r t i a r y - b u t y l ether;
every I I~ rise in temperature RON required increases by 1.5 TBA: t e r t i a r y - b u t y l alcohol; TAME: t e r t i a r y - a m y l - m e t h y l ether.
points [63]. Improving the octane number of fuel would result Source: Ref. [24],
in reducing power loss of the engine, improving fuel economy,
and a reduction in environmental pollutants and engine dam-
age. For these reasons, octane number is one of the important after addition of an additive. ONoxis the corresponding octane
properties related to the quality of gasolines. There are a num- number of oxygenate, which can be taken as the average val-
ber of additives that can improve octane number of gasoline ues for the ranges of RON and MON as given in Table 3.28. For
or jet fuels. These additives are tetra-ethyl lead (TEL), alco- example for MTBE, the range of RONox is 113-117; therefore,
hols, and ethers such as ethanol, methyl-tertiary-butyl ether for this oxygenate the value of RONox for use in Eq. (3.136)
(MTBE), ethyl-tertiary-butyl ether (ETBE), or tertiary-amyl is 115. Similarly the value MONox for this oxygenate is are
methyl ether (TAME). Use of lead in fuels is prohibited in 98. Equation (3.136) represents a simple linear relation for
nearly all industrialized countries due to its hazardous nature octane number blending without considering the interaction
in the environment, but is still being used in many third world between the components. This relation is valid for addition
and underdeveloped countries. For a fuel with octane number of additives in small quantities (low values of Xox,i.e., < 0.15).
(ON) of 100, increase in the ON depends on the concentra- However, when large quantities of two components are added
tion of TEL added. The following correlations are developed (i.e., two types of gasolines on 25:75 volume basis), linear mix-
based on the data provided by Speight [7]: ing rule as given by Eq. (3.136) is not valid and the interac-
tion between components should be taken into account [61 ].
Du Pont has introduced interaction parameters between two
TEL = -871.05 +2507.81 ~ - 2415.94 \ 1 0 0 / or three components for blending indexes of octane number
(oN 3 which are presented in graphical forms [89]. Several other
(3.134) +779.12 \ 1 0 0 ] blending approaches are provided in the literature [61]. The
simplest form of their tabulated blending indexes have been
ON = 100.35 + 11.06(TEL) - 3.406(TEL) 2 converted into the following analytical relations:
(3.135) + 0.577(TEL) 3 - 0.038(TEL) 4 BIRoN =
2
36.01+38.33X-99,8X +341.3X 3 -507.2X 4
+268.64X 5
11_<RON<76
where ON is the octane number and TEL is milliliter TEL - 2 9 9 . 5 + 1272X - 1552.9X 2 + 651X 3 76 _< RON _< 103
added to one U.S. gallon of fuel. These relations nearly repro-
2 2 0 6 . 3 - 4 3 1 3 . 6 4 X + 2178.57X 2 103 _< RON < 106
duce the exact data given by Speight and valid for ON above
/ x = RON/IO0
100. In these equations when clear octane number (without
TEL) is 100, TEL concentration is zero. By subtracting the cal- (3.137)
culated ON from 100, the increase in the octane number due
to the addition of TEL can be estimated, which may be used where BIRoNis the blending index for RON and should be used
to calculate the increase in ON of fuels with clear ON different together with Eq. (3.117) to calculate RON of a blend. Equa-
from 100. Equation (3.134) is useful to calculate amount of tion (3.137) reproduce the tabulated values of RON blending
TEL required for a certain ON while Eq. (3.135) gives ON of indexes with AAD of 0.06%.
fuel after a certain amount of TEL is added. For example, if Estimation of octane number of a fuel from its bulk proper-
0.3 mL of TEL is added to each U.S. gallon of a gasoline with ties is a challenging task, since ON very much depends on the
RON of 95, Eq. (3.135) gives ON of 104.4, which indicates an chemical structure of components of the mixture. Figure 3.32
increase of 4.4 in the ON. This increase is based on the refer- shows variation of RON with boiling point of pure hydrocar-
ence ON of 100 which can be used for ON different from 100. bons from different families as produced from data given in
Therefore, the ON of gasoline in this example will be 95 + 4.4 Table 2.2. If PIONA composition of a fuel is known, RON of a
or 99.4. Different relations for octane number of various fuels fuel may be estimated from the pseudocomponent techniques
(naphthas, gasolines, and reformates) in terms of TEL con- in the following form:
centration are given elsewhere [88]. RON = x~p(RON)Np + xI~(RON)Ip + xo(RON)o
Octane numbers of some oxygenates (alcohols and ethers)
(3.138) + xN(RON)N + XA(RON)A
are given in Table 3.28 [24]. Once these oxygenates are added
to a fuel with volume fraction of :Coxthe octane number of where x is the volume fraction of different hydrocarbon
product blend is [24] families i.e., n-paraffins (NP), isoparaffins (IP), olefins (O),
naphthenes (N), and aromatics (A). RONNp, RONIp, RONo,
(3.136) ON = Xox(ON)ox+ (1 - Xox)(ON)clear
RONN, and RONA are the values of RON of pseudocompo-
where ONclear is the clear octane number (RON or MON) of nents from n-paraffin, isoparaffins, olefins, naphthenes, and
a fuel and ON is the corresponding octane number of blend aromatics families whose boiling points are the same as the
140 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

8o
,2o ; 7 - : : .......... ...... ..................
" --omZ; 5l ! i I I I

,oo .

60 i ' - ' I , , o~ ,

.~ 40 n-paraft~s \ k~ \ l : ~ ~phthenes ~ so ,

~ 40 .

0 An-alkyIcyclopentanes : :x
On-alkylbenzenes
-20 . . . . . ' . . . . ' . . . .
-100 -50 0
' ....
50
i ....
100
,
150
,i;
200
$03~t J'"lol'oox:
"<~176r'""
BoilingPoint,~ j I . 1
II.3 11.4 1T.5 11.6 11.7 11.8 11,9 12.0 12.1 12.2 12.3
FIG. 3.32--Research octane number of different Characterization Factor
families of hydrocarbons. Taken with permission
from Ref. [68], FIG. 3.33--Research octane number of naphthas (~ = 1.8 x
~ +32). Taken with permission from Ref. [79].

m i d boiling p o i n t or the ASTM D 86 t e m p e r a t u r e at 50%

p o i n t of the fraction a n d can be d e t e r m i n e d from Fig. 3.32
o r Table 2.2. Generally p e t r o l e u m p r o d u c t s are free of olefins
a n d the m a i n groups p r e s e n t in a p e t r o l e u m p r o d u c t s are
n-paraffins, isoparaffins, n a p h t h e n e s , a n d aromatics. The role
of isoparaffins on octane n u m b e r is significant as they have
ON values greater t h a n n-paraffins. I n a d d i t i o n different types
of isoparaffins have different octane n u m b e r s at the s a m e
boiling point. As the n u m b e r of b r a n c h e s in an iso-paraffin
c o m p o u n d increases the octane n u m b e r also increases. F o r
this r e a s o n it w o u l d be m o r e a p p r o p r i a t e if RONIp in Eq.
(3.138) is an average value of octane n u m b e r s of various types
of isoparaffins. F o r convenience a n d c o m p u t e r calculations,
values of RON for these various h o m o l o g o u s h y d r o c a r b o n
g r o u p s have been c o r r e l a t e d to n o r m a l boiling point, Tb in
the following form:
(3.139) RON = a + b T +cT2 +dT3 + e T 4
w h e r e RON is the clear r e s e a r c h o c t a n e n u m b e r a n d T =
(Tb--273.15)/100 in w h i c h Tb is the boiling p o i n t in
kelvin. Based on the d a t a t a k e n from the API-TDB [2],
the coefficients a - e were d e t e r m i n e d a n d are given in
Table 3.29 [68, 78]. It s h o u l d be n o t e d that for isoparaf-
fins the coefficients are given for four different groups of
2-methylpentanes, 3-methylpentanes, 2,2-dimethylpentanes,
a n d 2,3-dimethylpentanes. Octane n u m b e r s of various
isoparaffins vary significantly a n d for this r e a s o n an average
value of RON for these four different iso-paraffinic groups is
c o n s i d e r e d as the value of RONIp for use in Eq. (3.138).
N o r m a l l y w h e n detailed PIONA c o m p o s i t i o n is n o t avail-
able, PNA c o m p o s i t i o n is p r e d i c t e d from the m e t h o d s pre-
sented in Section 3.5.1. F o r such cases Eq. (3.138) m a y he
simplified by considering xo = 0 a n d XNp = Xip = xp/2. Be-
cause RON of n-paraffins a n d isoparaffins differs significantly
(Fig. 3.32), the a s s u m p t i o n of equal a m o u n t s of n-paraffins
a n d isoparaffins can lead to substantial errors in calculation
of RON for fuels whose n o r m a l a n d iso paraffins contents dif-
fer significantly. F o r such cases this m e t h o d estimates RON of
a fuel w i t h a higher e r r o r b u t requires m i n i m u m i n f o r m a t i o n
on distillation a n d specific gravity.
Nelson [79] gives graphical relation for e s t i m a t i o n of RON
of n a p h t h a s in t e r m s of Kw c h a r a c t e r i z a t i o n factor o r paraffin
c o n t e n t (wt%) a n d m i d boiling p o i n t as given in Figs. 3.33
a n d 3.34, respectively.
As m e n t i o n e d earlier if a m o u n t of paraffins in wt% is n o t
available, vol% m a y b e used i n s t e a d of wt% if necessary. Once
RON is d e t e r m i n e d , MON c a n be calculated f r o m the follow-
ing relation p r o p o s e d by Jenkins [80]:
MON = 22.5 + 0.83 RON - 20.0 SG - 0.12 ( % 0 )
(3.140) + 0.5 ( T M L ) + 0.2 (TEL)
w h e r e SG is the specific gravity, TML a n d TEL are the con-
centrations of tetra methyl lead a n d tetra ethyl l e a d in mL/UK
gallon, a n d % 0 is the vol% of olefins in the gasoline. F o r olefin-
a n d lead-free fuels ( % 0 = TML = TEL = 0) a n d Eq. (3.140)
reduces to a simple form in t e r m s of RON a n d SG. F r o m this

TABLE 3.29--Coefficients for Eq. (3.139) for estimation of RON [68, 78].
Hydrocarbon family a b c d
n-Paraffins 92.809 -70.97 -53 20 10
isoparaffins
2-Methyl-pentanes 95.927 -157.53 561 -600 200
3-Methyl-pentanes 92.069 57.63 -65 0 0
2,2-Dimethyl-pentanes 109.38 -38.83 -26 0 0
2,3-Dirnethyl-pentanes 97.652 -20.8 58 -200 100
Naphthenes -77.536 471.59 -418 100 0
Aromatics 145.668 -54.336 16.276 0 0
Taken with permission from Ref. [68].
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS 141

9 i i ! b. To p r e d i c t PNA, we calculate M f r o m Eq. (2.50) as M =

79.54. Since M < 200 a n d viscosity is not available we use
Eqs. (3.77) a n d (3.78) a n d (3.72) to p r e d i c t the c o m p o s i -
tion. F r o m Eqs. (2.126) a n d (2.127), n20 = 1.3642 a n d f r o m
Eq. (3.50), m = - 8 . 8 1 5 9 . The p r e d i c t e d c o m p o s i t i o n for
6C ~tm
%P, %N, a n d %A is 95.4%, 7.4%, a n d - 2 . 8 % , respectively.
Since p r e d i c t e d %A is negative, it is set equal to zero a n d
Z 50 " '
, "175~ P"
the n o r m a l i z e d c o m p o s i t i o n is xp -- 0.928, xN -- 0.072, a n d
XA = 0.0. To use Eq. (3.139) we split the paraffin c o n t e n t
%' 200~ F.

rr
4~
30
J\\" \ .-
equally b e t w e e n n-paraffins a n d isoparaffins as xNp = Xip =
0.928/2 = 0.464. I n this case RON = 72.36. The e r r o r o n
calculation of RON is 2.76.
c. To use Fig. 3.33 we need Tb = 42.8~ -- 109~ a n d Kw,
\ , w h i c h from Eq. (2.13) is c a l c u l a t e d as Kw -- 12.75. Since
20 I
ii \ Xx \~ !oo~F.
the Kw is outside the r a n g e of values in Fig. 3.33, a c c u r a t e
I r e a d i n g is n o t possible, b u t f r o m value of the boiling p o i n t
10 I i \"?50 ~ P. I it is obvious t h a t the RON from e x t r a p o l a t i o n of the curves
10 20 30 40 50 60 70 80 90 100 is above 70.
Wt% Paraffin Hydrocarbons d. To use Fig. 3.34 we need total paraffins w h i c h is % - -
49.33 + 41.45 = 90.78 a n d Tb = 109~ In this case Tb is out-
FIG. 3.34--Fiesearch octane number versus paraffin content side the range of values on the curves, b u t with extrapola-
(~ ---- 1 . 8 x ~ +32). Taken with permission from Fief. [7g]. tion a value of a b o u t 66 can be read. The e r r o r is a b o u t
-3.6.
e q u a t i o n sensitivity of gasoline can be d e t e r m i n e d . Gasolines e. To calculate MON we use Eq. (3.140) with R O N =
with lower sensitivity are desirable. 69.6, SG = 0.6501, a n d % 0 -- TML = TEL -----0. The esti-
m a t e d value is (MON)e~t. = 67.3, w h i c h is in g o o d agree-
Example 3.22~A n a p h t h a s a m p l e f r o m a n Australian c r u d e m e n t w i t h the r e p o r t e d value of 66.2 [46] w i t h e r r o r of + 1.1.
oil has the following characteristics: boiling p o i n t r a n g e f. If e s t i m a t e d RON value of 70.55 (from P a r t a) is u s e d in
15.5-70~ specific gravity 0.6501, n-paraffins 49.33%, Eq. (3.140), the p r e d i c t e d MON is 68 with deviation of + 1.8.
isoparaffins 41.45%, n a p h t h e n e s 9.14%, a r o m a t i c s 0.08%, g. E r r o r s are c a l c u l a t e d a n d given in each part. E q u a t i o n
clear RON 69.6, a n d MON 66.2 [Ref. [46], p. 359). (3.138) gives the lowest e r r o r w i t h deviation of less t h a n 1
w h e n e x p e r i m e n t a l PIONA c o m p o s i t i o n is used. F o r sam-
a. E s t i m a t e RON f r o m the p s e u d o c o m p o n e n t m e t h o d using ples in w h i c h the difference b e t w e e n a m o u n t s of n-paraffins
experimental composition. a n d isoparaffins is small, Eq. (3.138) gives even b e t t e r re-
b. E s t i m a t e RON f r o m the p s e u d o c o m p o n e n t m e t h o d using sults. I n the cases t h a t the c o m p o s i t i o n is n o t available the
p r e d i c t e d PNA c o m p o s i t i o n . p r o c e d u r e used in P a r t b using p r e d i c t e d c o m p o s i t i o n w i t h
c. E s t i m a t e RON from Fig. 3.33. m i n i m u m d a t a on boiling p o i n t a n d specific gravity gives
d. E s t i m a t e RON f r o m Fig. 3.34. an a c c e p t a b l e value for RON. *
e. E s t i m a t e MON f r o m actual r e p o r t e d RON.
f. E s t i m a t e MON f r o m p r e d i c t e d RON from Part a.
g. F o r each case calculate the e r r o r (deviation between esti-
3.6.8 Carbon Residue
m a t e d a n d r e p o r t e d values).
W h e n a p e t r o l e u m fraction is v a p o r i z e d in the a b s e n c e of air
Solution--For this fraction: Tb = (15.5 + 70)/2 = 42.8~ at a t m o s p h e r i c pressure, the nonvolatile c o m p o u n d s have a
SG = 0.6501,Xp = 0.4933, xw = 0.4145, XN = 0.0914, XA : c a r b o n a c e o u s residue k n o w n as carbon residue, w h i c h is des-
0.008. i g n a t e d b y CR. Therefore, heavier fractions w i t h m o r e aro-
m a t i c contents have h i g h e r c a r b o n residues while volatile a n d
a. RON can be e s t i m a t e d f r o m Eq. (3.138) t h r o u g h pseu-
light fractions such as n a p h t h a s a n d gasolines have n o car-
d o c o m p o n e n t m e t h o d using R O N values for p u r e hydro-
b o n residues. CR is p a r t i c u l a r l y a n i m p o r t a n t c h a r a c t e r i s t i c
c a r b o n s calculated f r o m Eq. (3.139) a n d Table 3.29
of c r u d e oils a n d p e t r o l e u m residues. H i g h e r CR values in-
with Tb = 315.9 K. Results of calculation are (RON)np =
dicate low-quality fuel a n d less h y d r o g e n content. There are
54.63, (RON)Ip = (90.94 + 104.83 + 88 + 87.05)/4 = 92.7,
two o l d e r different test m e t h o d s to m e a s u r e c a r b o n residues,
(RON)N -- 55.57, a n d (RON)A ----125.39. I n calculation of
R a m s b o t t o m (ASTM D 524) a n d the C o n r a d s o n (ASTM
(RON)w, an average value for RON of four families in
D 189). The r e l a t i o n s h i p between these m e t h o d s are also given
Table 3.29 is calculated. F r o m Eq. (3.139), clear RON can
b y the ASTM D 189 m e t h o d . Oils that have ash f o r m i n g c o m -
be calculated as:
p o u n d s have e r r o n e o u s l y high c a r b o n residues b y b o t h meth-
RON = 0.4933 • 54.63 + 0.4145 x 92.7 + 0.0914 x 55.57 ods. F o r such oils ash should be r e m o v e d before the m e a s u r e -
+ 0 . 0 0 0 8 • 125.39 = 70.55. ment. There is a m o r e recent test m e t h o d (ASTM D 4530) t h a t
requires s m a l l e r s a m p l e a m o u n t s a n d is often referred as mi-
In c o m p a r i s o n with the r e p o r t e d value of 69.6 the e r r o r is crocarbon residue (MCR) a n d as a result it is less precise in
70.55 - 69.6 = 0.95. practical t e c h n i q u e [7]. In m o s t cases c a r b o n residues are
142 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
reported in wt% by Conradson method, which is designated
by %CCR.
Carbon residue can be correlated to a number of other prop-
erties. It increases with an increase in carbon-to-hydrogen
ratio (CH), sulfur content, nitrogen content, asphaltenes con-
tent, or viscosity of the oil. The most precise relation is
between CR and hydrogen content in which as hydrogen con-
tent increases the carbon residues decreases [7]. The hydro-
gen content is expressed in terms of H/C atomic ratio and
the following relation may be used to estimate CCR from H/C
[81].
(3.141) %CCR = 148.7 - 86.96 H/C
if H/C _> 1.71 (%CCR < 0), set %CCR = 0.0 and if H/C < 0.5
(%CCR > 100), set %CCR = 100. H/C ratio can be estimated
from CH ratio methods given in Section 2.6.3.
The carbon residue is nearly a direct function of high boil-
ing asphaltic materials and Nelson has reported a linear re-
lation between carbon residue and asphalt yield [82]. One
of the main characteristic of residuum is its asphaltene con-
tent. Asphaltenes are insoluble in low molecular weight n-
alkanes including n-pentane. Knowledge of n-pentane insol-
ubles in residual oils is quite important in determining yields
and products qualities for deasphalting, thermal visbreaking,
and hydrodesulfurization processing. The relation between
the normal pentane insolubles and carbon residue is as fol-
lows [61]:
(3.142) %NCs = 0.74195 (%CCR) + 0.01272 (%CCR) 2
where %NC5 is the wt% of n-pentane insolubles and %CCR is
the wt% of Conradson carbon residue. Once %NCs is known,
the asphaltene content (asphaltene wt%) of a residue can be
determined from the following empirical relation:
(3.143) %Asphaltene = a(%NCs)
where a is 0.385 for atmospheric residue and 0.455 for vac-
uum residues [61,66]. These equations are approximate and
do not provide accurate predictions.
Example 3.23--A vacuum residue of an Australian crude oil
has carbon-to-hydrogen weight ratio of 7.83. Estimate its car-
bon residue and asphaltene contents and compare the results
with the experimental values of 15.1 and 4.6%, respectively
[46].
Solution--With CH=7.83, from Eq. (2.122), HC atomic ratio
is calculated as HC --- 1.52. From Eq. (3.141), %CCR = 16.4%
and from Eq. (3.142), %NC5 = 15.6%. From Eq. (3.143) with
a = 0.455 (for vacuum residue) we calculate %Asphaltene =
7.1. The results show that while Eq. (3.141) provides a
good prediction for %CCR, prediction of %Asphaltene from
Eq. (3.143) is approximate.
3.6.9 S m o k e P o i n t
Smoke point is a characteristic of aviation turbine fuels and
kerosenes and indicates the tendency of a fuel to burn with
a smoky flame. Higher amount of aromatics in a fuel causes
a smoky characteristic for the flame and energy loss due to
thermal radiation. The smoke point (SP) is a maximum flame
height at which a fuel can be burned in a standard wick-fed
lamp without smoking. It is expressed in millimeters and a
high smoke point indicates a fuel with low smoke-producing
tendency [61]. Measurement of smoke point is described un-
der ASTM D t322 (U.S.) or IP 57 (UK) and ISO 3014 test
methods. For a same fuel measured smoke point by IP test
method is higher than ASTM method by 0.5-1 m m for smoke
point values in the range of 20-30 m m [61].
Smoke point may be estimated from either the PNA com-
position or from the aniline point. The SP of kerosenes from
IP test method may be estimated from the following relation
[90]:
SP = 1.65X - 0.0112X 2 - 8.7
(3.144) 100
X =
0.61Xp + 3.392xN + 13.518xA
where SP is the smoke point by IP test method in m m and Xv,
xN, and :cA are the fraction of paraffin, naphthene, and aro-
matic content of kerosenes. The second method is proposed
by Jenkins and Walsh as follows [83]:
SP = -255.26 + 2.04AP - 240.8 ln(SG) + 7727(SG/AP)
(3.145)
where AP is the aniline point in ~ and SG is the specific grav-
ity at 15.5~ Both Eqs. (3.144) and (3.145) estimate SP ac-
cording to the IP test method. To estimate SP from the ASTM
D 1322 test method, 0.7 mm should be subtracted from the
calculated IP smoke point. Equations (3.144) and (3.145) are
based on data with specific gravity in the range of 0.76-0.82,
and smoke points in the range of 17-39 mm. Based on some
preliminary evaluations, Eq. (3.133) is expected to perform
better than Eq. (3.144), because smoke point is very much re-
lated to the aromatic content of the fuel which is expressed in
terms of aniline point in the Jenkins-Walsh method. In addi-
tion the specific gravity, which is an indication of molecular
type, is also used in the equation. Equation (3.144) may be
used for cases that the aniline point is not available but ex-
perimental PNA composition is available. Albahri et al [68]
also proposed the following relation for prediction of smoke
point using API gravity and boiling point:
(3.146) SP = 0.839(API) + 0.0182634(Tb) -- 22.97
where SP is in m m (ASTM method) and Tb is the average
boiling point in kelvin. This equation when tested for 136
petroleum fractions gave an average error of about 2 m m [68].
Example 3.24--A Nigerian kerosene has an API gravity of
41.2, aniline point of 55.6~ and the PNA composition of
36.4, 49.3, and 14.3%. Estimate the smoke point of this fuel
from. Equations (3.144)-(3.146) and compare with the exper-
imental value of 20 mm (Ref. [46], p. 342).
Solution--To estimate SP from Eq. (3.144), we have Xp =
0.364, XN = 0.493, and XA= 0.143 which give X = 26.13. Cal-
culated SP is 26.8 m m according to the IP method or 26.1
mm according to the ASTM method. To use Eq. (3.145) we
have from API gravity, SG = 0.819, AP -- 55.6~ the calcu-
lated SP is SP = 20 mm. The ASTM smoke point is then 19.3
m m which is in very good agreement with the experimental
value of 20 with deviation of -0.7 mm. Predicted value from
Eq. (3.146) is 17.6 mm.
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS 143

3.7 QUALITY OF P E T R O L E U M P R O D U C T S % Reduction

M e t h o d s p r e s e n t e d in this c h a p t e r c a n be u s e d to evaluate
the quality of p e t r o l e u m p r o d u c t s f r o m available p a r a m e t e r s .
The quality of a p e t r o l e u m p r o d u c t d e p e n d s on certain -5]
specifications o r p r o p e r t i e s of the fuel to satisfy r e q u i r e d
criteria set b y the m a r k e t d e m a n d . These characteristics are
specified for b e s t use of a fuel (i.e., highest engine perfor-
m a n c e ) or for cleaner e n v i r o n m e n t while the fuel is in use.
These specifications vary f r o m one p r o d u c t to a n o t h e r a n d -10
from one c o u n t r y to another. F o r example, for gasoline the
quality is d e t e r m i n e d b y a series of p r o p e r t i e s such as sulfur
a n d a r o m a t i c contents, octane number, v a p o r pressure,
h y d r o g e n content, a n d boiling range. Engine w a r m - u p t i m e
is affected by the p e r c e n t distilled at 70~ a n d the ASTM 90% -15
t e m p e r a t u r e . F o r the a m b i e n t t e m p e r a t u r e of 26.7~ (80~
a gasoline m u s t have ASTM 90% t e m p e r a t u r e of 188~ a n d
3% distilled at 70~ to give a c c e p t a b l e w a r m - u p t i m e [63].
S t a n d a r d o r g a n i z a t i o n s such as ASTM give such specifica-
-20
tions for various products. F o r r e f o r m u l a t e d gasoline sulfur f
HC CO NOx
c o n t e n t of less t h a n 300 p p m (0.03 wt%) is r e q u i r e d [63].
A m o u n t of p a r t i c u l a t e emissions is directly related to the FIG. 3.35--Influence of sulfur content in gasoline (from
a r o m a t i c a n d sulfur content of a fuel. Figure 3.35 shows the 500 to 50 ppm) in reduction of pollutant gases. Taken with
influence of sulfur r e d u c t i o n in gasoline f r o m 500 to 50 p p m permission from Ref. [24].
in the r e d u c t i o n of p o l l u t a n t emissions [24].
Vapor p r e s s u r e of gasoline of jet fuel d e t e r m i n e s their ig- p r o d u c e the r e q u i r e d final product. In E x a m p l e 3.20, it was
nition characteristics. While freezing p o i n t is i m p o r t a n t for s h o w n t h a t to have a p r o d u c t with certain flash point, one c a n
jet fuels it is n o t a m a j o r c h a r a c t e r i s t i c for gas oils. F o r lubri- d e t e r m i n e the v o l u m e of various c o m p o n e n t s in the blend.
cating oil p r o p e r t i e s such as viscosity a n d viscosity index are The s a m e a p p r o a c h c a n be e x t e n d e d to a n y o t h e r property. F o r
i m p o r t a n t in a d d i t i o n to sulfur a n d PNA c o m p o s i t i o n . Aniline example, to increase v a p o r p r e s s u r e of gasoline n-butane m a y
p o i n t is a useful characteristic to indicate p o w e r of solubil- be a d d e d d u r i n g w i n t e r season to i m p r o v e engine starting
ity of solvents as well as a r o m a t i c contents of certain fuels. characteristics of the fuel [63]. The a m o u n t of r e q u i r e d b u t a n e
F o r heavy p e t r o l e u m p r o d u c t s knowledge of p r o p e r t i e s such to r e a c h a certain v a p o r p r e s s u r e value can be d e t e r m i n e d
as c a r b o n residue, p o u r point, and cloud p o i n t are of inter- t h r o u g h calculation of v a p o r p r e s s u r e b l e n d i n g index for the
est. S o m e i m p o r t a n t specifications of jet fuels are given in c o m p o n e n t s a n d the p r o d u c t as discussed in Section 3.6.1.1.
Table 3.30.
One of the techniques u s e d in refining t e c h n o l o g y to pro-
duce a p e t r o l e u m p r o d u c t w i t h a certain characteristic is the 3.8 M I N I M U M LABORATORY DATA
b l e n d i n g m e t h o d . Once a certain value for a p r o p e r t y (i.e.,
viscosity, octane number, p o u r point, etc.) of a p e t r o l e u m As discussed earlier m e a s u r e m e n t of all p r o p e r t i e s of various
p r o d u c t is required, the mixture m a y be b l e n d e d with a cer- p e t r o l e u m fractions a n d p r o d u c t s in the l a b o r a t o r y is an
tain c o m p o n e n t , additive o r a n o t h e r p e t r o l e u m fraction to i m p o s s i b l e task due to the r e q u i r e d cost a n d time. However,

TABLE 3.30---Some general characteristics of three fuels [24].

Specificationsa
Characteristics Gasolineb Jet fuel Diesel fuel
Max. total sulfur, wt% 0.05 0.2 0.05
Max aromatics content, vol% 20
Max olefins content, vol% 5
Distillation at 10 vol%,~ 204
Max final boiling point, ~ 215 300 370 c
Range of % evaporation at 70~ (E70) 15-47
Min research octane number (RON) 95
Min flash point, ~ 38 55
Specific gravity range 0.7254).78 0.775-0.84 0.82-0.86
Min smoke point, ~ 25
Max freezing point, ~ -47
Range of Reid vapor pressure, bar 0.35-0.90 d
Min cetane number 49
Taken with permission from Ref. [24].
aEuropean standards in the mid 1990s.
bEuropean unleaded Super 98 premium gasoline.
CAtthis temperature minimum of 95 vol% should be evaporated.
dVarieswith season according to the class of gasoline.
144 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

there are a n u m b e r of basic p a r a m e t e r s that m u s t be k n o w n boiling p o i n t is available, nearly all other p a r a m e t e r s c a n be

for a fraction to determine various properties from the
methods presented i n this chapter. As m o r e experimental
data are available for fraction a better characterization of
the fraction is possible. For example, to estimate sulfur
c o n t e n t of a fraction from Eqs. (3.97) a n d (3.98), the i n p u t
p a r a m e t e r s of specific gravity, molecular weight, density a n d
refractive index at 20~ are needed. If experimental values of
all these p a r a m e t e r s are available a good estimate of sulfur
c o n t e n t can be obtained. However, since n o r m a l l y all these
data are not available, M, n, a n d dE0 should be estimated from
SG a n d Tb. Therefore, a m i n i m u m of two p a r a m e t e r s that
are boiling p o i n t a n d specific gravity are needed to estimate
the sulfur content. However, for heavy fractions i n which
distillation data are not reported, M should be estimated
from k i n e m a t i c viscosity at 38 a n d 99~ (1)38 a n d 1)99) a n d
specific gravity t h r o u g h Eq. (2.52). Once M is estimated,
n c a n be estimated from M a n d SG t h r o u g h Eq. (2.127) a n d
d is calculated from SG t h r o u g h Eq. (2.123). With the knowl-
edge of M a n d SG all other p a r a m e t e r s can be estimated from
m e t h o d s presented in Chapter 2. Therefore, at least three
parameters of I)38, v99, a n d SG m u s t be k n o w n to d e t e r m i n e
sulfur c o n t e n t or other characteristics. I n a case that ouly one
viscosity data is known, i.e., vas, k i n e m a t i c viscosity at 99~
1)99, can be estimated from Eq. (2.61). I n this way estimated
value of M is less accurate t h a n the case that three values of
1)38, I)99 a n d SG are k n o w n from experimental m e a s u r e m e n t s .
We see that again for heavy fractions with knowledge of only
two parameters (i.e., 1)3a a n d SG or 1)99 a n d SG) all basic
properties of the fraction can be estimated. Therefore, to
o b t a i n the basic characterization parameters of a p e t r o l e u m
fraction a m i n i m u m of two parameters are needed.
If the only i n f o r m a t i o n is the distillation curve, t h e n spe-
cific gravity can be estimated from T10 a n d T50 t h r o u g h Eq.
(3.17) a n d Table 3.4. Having T50 a n d SG, all other p a r a m e t e r s
can be estimated as discussed above. W h e n only a p o r t i o n of
distillation curve (i.e., T2o, T40, a n d T60) is available, t h r o u g h
Eq. (3.35) the complete curve can be predicted a n d from this
equation T10 a n d Ts0 can be determined. Therefore, a p o r t i o n
of distillation curve can also be used to generate all p a r a m e -
ters related to properties a n d quality of petroleum fractions.
We showed that with the knowledge of PNA c o m p o s i t i o n a
better characterization of a fraction is possible t h r o u g h pseu-
d o c o m p o n e n t technique. Therefore, if the composition along
d e t e r m i n e d t h r o u g h m i d boiling p o i n t a n d PNA c o m p o s i t i o n
with better accuracy t h a n u s i n g only Tb a n d SG. For heavy
fractions in which Tb m a y not be available, the pseudocom-
p o n e n t technique c a n be applied t h r o u g h use of M a n d PNA
c o m p o s i t i o n where M m a y be estimated from viscosity data
if it is n o t available. As there are m a n y scenarios to estimate
basic properties of p e t r o l e u m fractions, use of available data
to predict the most accurate characterization p a r a m e t e r s is
a n engineering art which has a direct i m p a c t o n s u b s e q u e n t
prediction of physical properties a n d eventually on design cal-
culations. The basic laboratory data that are useful in charac-
terization m e t h o d s based o n their significance a n d simplicity
are given below:
1. distillation data, boiling p o i n t
2. specific gravity
3. c o m p o s i t i o n (i.e, PNA content)
4. molecular weight
5. refractive index
6. elemental analysis (i.e., CHS composition)
7. k i n e m a t i c viscosity at 37.8 a n d 98.9~ (100 a n d 210~
One can best characterize a p e t r o l e u m fraction if all the
above p a r a m e t e r s are k n o w n from laboratory m e a s u r e m e n t s .
However, a m o n g these seven items at least two items m u s t be
k n o w n for characterization purposes. I n any case w h e n exper-
i m e n t a l value for a characterization p a r a m e t e r is available it
should be used instead of predicted value. Among these seven
items that can be m e a s u r e d i n laboratory, refractive index a n d
specific gravity are the most c o n v e n i e n t properties to mea-
sure. Molecular weight especially for heavy fractions is also
very useful to predict other properties. As discussed i n Chap-
ter 2, for light fractions (M < 300; Nc < 22, Tb < 350~ the
best two pairs of parameters in the order of their character-
izing power are (Tb, SG), (Tb, n), (M, SG), (M, n), (v, SG), (Tb,
CH), (M, CH), (v, CH). The most suitable pair is (Tb, SG) a n d
the least one is (v, CH). As it is explained i n the next section,
for heavy fractions three p a r a m e t e r correlations are more ac-
curate. Therefore, for heavy fractions in which boiling p o i n t
c a n n o t be m e a s u r e d a m i n i m u m of three p a r a m e t e r s such
as viscosity at two different temperatures a n d specific gravity
(i.e., v38, I)99, SG) are needed. For heavy fractions the pseudo-
c o m p o n e n t m e t h o d is m u c h more accurate t h a n use of b u l k
properties for the e s t i m a t i o n of various properties. Therefore,

TABLE 3.31---Standard test methods for measurement of some properties of liquid petroleum products.
Property ASTMD IP ISO Property ASTMD IP ISO
Aniline Point 611 2/98 2977 Flash Point 93 34/97 2719
Carbon Residue (Ramsbottom) 524 14/94 4262 Freezing Point 2386 16/98 3013
Carbon Residue (Conradson) 189 13/94 6615 Hydrocarbon Types 1319 156/95 3837
Centane Number 4737 380/98 4264 Heating Value 240 12
Cloud Point 2500 219/94 3016 Kinematic Viscosity 445 71/97 3104
Color 1500 1 9 6 / 9 7 2049 Octane Number (Motor) 2700 236 5163
Density/Sp. Gr. 4052 365/97 2185 Refractive Index 1218
Distillation at Atm. Pressure 86 1 2 3 / 9 9 " 3405 Pour Point 97 15/95 3015
Distillation at Reduced Pressures 1160 6616" Sulfur Content 1266 1 0 7 / 8 6 2192
Distillation by Gas Chromatography 2887 406/99* Thermal Conductivity 2717-95
Distillation of Crude Oils 2892 8708* Vapor Pressure (Reid) 323 69/94 3007
Viscosity (Viscous Oils) 2983 370/85*
ASTM has test methods for certain properties for which other test methods do not suggest equivalentprocedures. Some of these methods
include heat of combustion:D 4809; smoke point: D 1322;surface tension: D 3825;vapor-liquidratio: D 2533; viscositytemperature chart: D
341; autoignition:D 2155 (ISO 3988). Further test methods for some specificproperties are givenin the text where the property is discussed.
ASTMmethods are taken from Ref. [4]. IP methods are taken from Ref. [85]. Methods specifiedby * are similar but not identical to other
standard methods. Most IP methods are also used as British Standard under BS2000methods [85]. The number after IP indicatesthe year of
]ast approval. ISO methods are taken from Refs. [24] and [85].
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS
the knowledge of PNA composition for prediction of prop-
erties of heavy fractions is more useful than for light frac-
tions. For wide boiling range fractions knowledge of complete
distillation curve is quite useful to consider nature of different
compounds and their effects on the properties of the mixture.
As it is shown in Chapter 4, for wide and heavy fractions such
as C7+ fractions, distribution of carbon number in the frac-
tion is the most useful information besides specific gravity.
Further analysis of minimum laboratory data for characteri-
zation of petroleum fractions is provided in our previous work
[84].
Predictive methods of characterization must be used when
experimental data are not available. If possible, one should
make m a x i m u m use of available experimental data, A sum-
mary of standard test methods for some specifications of
liquid petroleum products is given in Table 3.31. For some
properties equivalent test methods according to the interna-
tional standards organization (ISO) are also specified in this
table [24, 85].
3.9 ANALYSIS OF LABORATORY DATA A N D
D E V E L O P M E N T OF P R E D I C T I V E M E T H O D S
In Chapter 2 and this chapter the predictive methods in terms
of readily available parameters are presented for estimation of
various properties related to basic characteristics and quality
of petroleum fractions. Generally these methods fall within
two categories of empirical and semiempirical correlations.
In an empirical correlation the structure of the correlation
is determined through fitting the data and the type of input
parameters in each correlation are determined through anal-
ysis of experimental data. While in a semi-empirical corre-
lation, the structure and functionality of the relation is de-
termined from a theoretical analysis of parameters involved
and through analysis of existing theoretical relations. Once
the main functionality and nature of a correlation between
various physical properties is determined, the correlation co-
efficients can be determined from experimental data. The best
example of such a predictive method is development of Eq.
(2.38) in Chapter 2, which was developed based on the un-
derstanding of the intermolecular forces in hydrocarbon sys-
tems. This generalized correlation has been successfully used
to develop predictive methods for a variety of physical prop-
erties. An example of an empirical correlation is Eq. (2.54)
developed for estimation of molecular weight of petroleum
fractions. Many other correlations presented in this chapter
for estimation of properties such as aniline and smoke points
or methods presented for calculation of octane numbers for a
blend are also purely empirical in nature. In development of
an empirical relation, knowledge of the nature of properties
involved in the correlation is necessary. For example, aniline
point is a characteristic that depends on the molecular type
of hydrocarbons in the fraction. Therefore, it is appropriate
to relate aniline point to the parameters that characterize hy-
drocarbon types (i.e., Ri) rather than boiling point that char-
acterizes carbon number in a hydrocarbon series.
Mathematical functions can be expressed in the form of
polynomial series; therefore, it is practically possible to de-
velop correlations in the forms of polynomial of various de-
grees. With powerful computational tools available at present
it is possible to find an empirical correlation for any set of
145
laboratory data for any physical property in terms of some
other parameters. As the complexity and the n u m b e r of pa-
rameters increases the accuracy of the correlation also in-
creases with respect to the data used in the development of
the correlation. However, the main problem with empirical
correlations is their limited power of extrapolation and the
large number of numerical constants involved in the correla-
tion. For example, in Chapter 2, several correlations are pro-
vided to estimate molecular weight of hydrocarbons in terms
of boiling point and specific gravity. Equation (2.50) derived
from Eq. (2.38) has only three numerical constants, which
are developed from molecular weight of pure hydrocarbons.
Tsonopoulos et al. [86] made an extensive analysis of various
methods of estimation of molecular weights of coal liquids.
Equation (2.50) was compared with several empirical corre-
lations specifically developed for coal liquids having as m a n y
as 16 numerical constants. They concluded that Eq. (2.50)
is the most accurate method for the estimation of molecular
weight of coal liquids. No data on coal liquid were used in
development of constants in Eq. (2.50). However, since it was
developed with some physical basis and properties of pure
hydrocarbons were used to obtain the numerical constants
the equation has a wide range of applications from pure hy-
drocarbons to petroleum fractions and coal liquids, which are
mainly aromatics. This indicates the significance of develop-
ment of correlations based on the physical understanding of
the nature of the system and its properties. The main advan-
tage of such correlations is their generality and simplicity.
The main characteristics of an ideal predictive method for a
certain property are accuracy, simplicity, generality, and avail-
ability of input parameters. The best approach toward the
development of such correlations would be to combine phys-
ical and theoretical fundamentals with some modifications.
An example of such type of correlations is Eq. (2.39), which is
an extension of Eq. (2.38) derived from physical basis. A pure
empirical correlation might be quite accurate to represent the
data used in its development but when it is applied to other
systems the accuracy is quite low. In addition characterizing
the systems according to their degree of complexity is helpful
to develop more accurate correlations. For example, heavy
fractions contain heavy and nonpolar compounds, which dif-
fer with low-molecular-weight hydrocarbons present in light
petroleum fractions. Therefore, in order to increase the de-
gree of accuracy of a predictive method for a certain prop-
erty it is quite appropriate to develop one correlation for
light and one correlation for heavy fractions. For heavy frac-
tions because of the nature of complex compounds in the
mixture three input parameters are required. As variation in
properties of pure hydrocarbons from one family to another
increases with increase in carbon numbers (i.e., see Figs. 2.15
and 2.23), the role of composition on estimation of such prop-
erties for heavy fractions is more than its effect on the prop-
erties of light fractions. Therefore, including molecular type
in the development of predictive methods for such properties
of heavy fractions is quite reasonable and useful and would
enhance the accuracy of the method.
Once the structure of a correlation is determined from
theoretical developments between various properties, ex-
perimental data should be used to determine the numerical
constants in the correlation. If the data on properties of
pure hydrocarbons from different families are used to
determine the constants, the resulting correlation would be
146 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
more general and applicable to various types of petroleum
fractions within the same boiling point range. The accuracy
of a correlation for a specific group of fractions could be
increased if the coefficients in the correlation are obtained
from the same group of fractions. In obtaining the constants,
many equations may be convertible into linear forms and
spreadsheets such as Lotus or Excel programs may be used to
obtain the constants by means of least squared method. Non-
linear regression of correlations through these spreadsheets
is also possible. In analyzing the suitability of a correlation,
the best criteria would be the R2 or correlation parameter
defined by Eq. (2.136) in which values of above 0.99 indicate
an equation is capable of correlating data. Use of a larger data
bank and most recent published data in obtaining the nu-
merical constants would enhance accuracy and applicability
of the correlation. A fair way of evaluations and comparison
of various correlations to estimate a certain property from
the same input parameters would be through a data set not
used in obtaining the coefficients in the correlations. In such
evaluations AAD or %AAD can be used as the criterion to
compare different methods. Average absolute deviation may
be used when the range of variation in the property is very
large and small values are estimated. For example, in predic-
tion of PNA composition, amount of aromatics varies from
1% in petroleum fractions to more than 90% in coal liquids.
AAD of 2 (in terms of percentage) in estimating aromatic
content is quite reasonable. This error corresponds to 200%
in terms of %AD for fractions with t% aromatic content.
Experience has shown that correlations that have fewer nu-
merical constants and are based on theoretical and physical
grounds with constants obtained from a wide range of data
set are more general and have higher power of extrapolation.
3.10 CONCLUSIONS AND
RECOMMENDATIONS
In this chapter various characterization methods for differ-
ent petroleum fractions and mixtures have been presented.
This is perhaps one of the most important chapters in the
book. As the method selected for characterization of a frac-
tion would affect prediction of various properties discussed in
the remaining part of the book. As it is discussed in Chapter 4,
characterization and estimation of properties of crude oils de-
pend on the characterization of petroleum fractions discussed
in this chapter. Through methods presented in this chapter
one can estimate basic input data needed for estimation of
thermodynamic and physical properties. These input param-
eters include critical properties, molecular weight, and acen-
tric factor. In addition methods of estimation of properties
related to the quality of a petroleum product such as distil-
lation curves, PNA composition, elemental composition, vis-
cosity index, carbon residue, flash, pour, cloud, smoke, and
freezing points as well as octane and cetane numbers are pre-
sented. Such methods can be used to determine the quality
of a fuel or a petroleum product based on the minimum labo-
ratory data available for a fraction. Methods of conversion of
various types of distillation curves help to determine neces-
sary information for process design on complete true boiling
point distillation curve when it is not available. In addition a
method is provided to determine complete distillation curve
based on a minimum of three data points along the distillation
curve. For prediction of each characteristic of a petroleum
fraction, several methods are provided that have been in use
in the petroleum industry. Limitations, advantages, and dis-
advantages of each method are discussed.
Basically two approaches are proposed in characterization
of petroleum fractions. One technique is based on the use
of bulk properties (i.e., Tb and SG) considering the whole
mixture as a single pseudocomponent. The second approach
called pseudocomponent technique considers the fraction as
a mixture of three pseudocomponents from the three fami-
lies. This technique is particularly useful for heavy fractions.
A third approach is also provided for wide boiling range frac-
tions. However, since behavior of such fractions is similar to
that of crude oils the technique is mainly presented in the next
chapter. Fractions are generally divided into light and heavy
fractions. For heavy fractions a minimum of three character-
ization parameters best describe the mixture. Recommenda-
tions on the use of various input parameters and advantages
of different methods were discussed in Sections 3.8 and 3.9.
For light fractions (M < 300, Nc < 22) and products of atmo-
spheric distillation unit, Eq. (2.38) for M, I, and d is quite ac-
curate. For such light fractions, To, Pc, and cocan be estimated
from Eqs. (2.65), (2.66), and (2.105), respectively. For pre-
diction of the PNA composition for fractions with M < 200,
Eqs. (3.77) and (3.78) in terms of m and SG are suitable. For
fractions with M > 200, Eqs. (3.71)-(3.74) in terms of P~ and
VGC are the most accurate relations; however, in cases that
viscosity data are not available, Eqs. (3.79) and (3.80) in terms
of P~ an CH are recommended. Special recommendations on
use of various correlations for estimation of different prop-
erties of petroleum fractions from their bulk properties have
been given in Section 2.10 and Table 2.16 in Chapter 2.
3.11 P R O B L E M S
3.1. List four different types of analytical tools used for com-
positional analysis of petroleum fractions.
3.2. What are the advantages/disadvantages and the differ-
ences between GC, MS, GC-MS, GPC and HPLC instru-
ments?
3.3. A jet naphtha has the following ASTM D 86 distillation
curve [ 1]:
vol% distilled 10 30 50 70 90
ASTM D 86 temperature, ~ 151.1 156.1 160.6 165.0 171.7
a. Calculate VABP, WABR MABP, CABP, and MeABP for
this fraction. Comment on your calculated MeABR
b. Estimate the specific gravity of this fraction and com-
pare with reported value of 0.8046.
c. Calculate the Kw for this fraction and compare with
reported value of 11.48.
3.4. A kerosene sample has the following ASTM D 86, TBP,
and SG distribution along distillation curve [1]. Convert
ASTM D 86 distillation curve to TBP by Riazi-Daubert
and Daubert's (API) methods. Draw actual TBP and pre-
dicted TBP curves on a single graph in ~ Calculate the
average specific gravity of fraction form SG distribution
and compare with reported value of 0.8086.
 3. C H A R A C T E R I Z A T I O N OF P E T R O L E U M F R A C T I O N S 147

Vol% IBP 5 10 20 30 40 50 60 70 80 90 95 FBP

ASTM, ~ 330 342 350 366 380 390 404 417 433 450 469 482 500
TBP, ~ 258 295 312 337 358 ... 402 ... 442 ... 486 499 ...
SG ... 0.772 0.778 0.787 0.795 ... 0.817 ... 0.824 ... 0.829 0.830 ...

3.5. F o r the k e r o s e n e s a m p l e of P r o b l e m 3.4, a s s u m e ASTM
t e m p e r a t u r e s at 30, 50, a n d 70% p o i n t s are the only
k n o w n i n f o r m a t i o n . Based on these d a t a points o n the
distillation curve p r e d i c t the entire distillation curve a n d
c o m p a r e with the a c t u a l values in b o t h t a b u l a t e d a n d
g r a p h i c a l forms.
3.6. F o r the kerosene s a m p l e of P r o b l e m 3.4 calculate the
W a t s o n K factor a n d generate the SG d i s t r i b u t i o n b y
c o n s i d e r i n g c o n s t a n t Kw along the curve. Calculate
%AAD for the deviations b e t w e e n p r e d i c t e d SG a n d ac-
tual SG for t h e 8 d a t a points. S h o w a g r a p h i c a l evalua-
tion of p r e d i c t e d distribution.
3.7. F o r the kerosene s a m p l e of P r o b l e m 3.4 generate SG
d i s t r i b u t i o n using Eq. (3.35) a n d d r a w p r e d i c t e d SG dis-
t r i b u t i o n w i t h actual values. Use Eq. (3.37) to calculate
specific gravity of the m i x t u r e a n d c o m p a r e with the ac-
tual value of 0.8086.
3.8. F o r the kerosene s a m p l e of P r o b l e m 3.4 calculate den-
sity at 75~ a n d c o m p a r e with the r e p o r t e d value o f
0.8019 g/cm 3 [1].
3.9. F o r the kerosene s a m p l e of P r o b l e m 3.4, a s s u m e the only
d a t a available are ASTM D 86 t e m p e r a t u r e s at 30, 50,
a n d 70% points. Based on these d a t a calculate the spe-
cific gravity at 10, 30, 50, 70, a n d 90 % a n d c o m p a r e w i t h
the values given in P r o b l e m 3.4.
3.10. A p e t r o l e u m fraction has the following ASTM D 1160
distillation curve at 1 m m H g :
vol% distilled 10 30 50 70 90
ASTMD 1160 temperature, ~ 104 143 174 202 244
Predict TBP, ASTM D 86, a n d EFV distillation curves all
at 760 m m Hg.
3.11. A gas oil s a m p l e has the following TBP a n d density dis-
t r i b u t i o n [32]. E x p e r i m e n t a l values of M, n20, a n d SG
are 214, 1.4694, a n d 0.8475, respectively.
a. Use the m e t h o d o u t l i n e d for wide boiling range frac-
tions a n d e s t i m a t e M a n d n20.
b. Use Eqs. (2.50) a n d (2.115) to e s t i m a t e M a n d n
c. C o m p a r e %AD f r o m m e t h o d s a a n d b.
3.13. F o r the fraction of P r o b l e m 3.12 calculate acentric fac-
tor, w from the following methods:
a. L e e - K e s l e r m e t h o d w i t h Tc a n d Pc f r o m Parts (a), (b),
(c), a n d (d) of P r o b l e m 3.12.
b. E d m i s t e r m e t h o d w i t h Tc a n d Pc from Parts a n d (a)
a n d (e) of P r o b l e m 3.12.
c. K o r s t e n m e t h o d with Tc a n d Pc f r o m P a r t (a) of Prob-
lem 3.12.
d. P s e u d o c o m p o n e n t m e t h o d
3.14. A p u r e h y d r o c a r b o n has a boiling p o i n t of 110.6~ a n d
a specific gravity of 0.8718. W h a t is the type of this hy-
d r o c a r b o n (P, N, MA, o r PA)? H o w can you check y o u r
answer? Can you guess the c o m p o u n d ? In y o u r analysis
it is a s s u m e d that you do not have access to the table of
p r o p e r t i e s of p u r e h y d r o c a r b o n s .
3.15. E x p e r i m e n t a l d a t a on sulfur content of s o m e p e t r o l e u m
p r o d u c t s along with o t h e r b a s i c p a r a m e t e r s are given in
Table 3.32. It is a s s u m e d that the only d a t a available for
a p e t r o l e u m p r o d u c t is its m i d boiling point, Tb, ( c o l u m n
1) a n d refractive index at 20~ n20 ( c o l u m n 2). Use ap-
p r o p r i a t e m e t h o d s to c o m p l e t e c o l u m n s 4,6,8,9,10, a n d
12 in this table.
3.16. ASTM 50% t e m p e r a t u r e (Tb), specific gravity (SG), a n d
the PNA c o m p o s i t i o n of 12 p e t r o l e u m fractions are given
in Table 3.33. Complete c o l u m n s of this table b y calculat-
ing M, n, m, a n d the PNA c o m p o s i t i o n for each fraction
b y using a p p r o p r i a t e methods.
3.17. A gasoline p r o d u c t from N o r t h Sea c r u d e oil has boiling
range of C5-85~ a n d specific gravity of 0.6771. Predict
the PNA c o m p o s i t i o n a n d c o m p a r e with e x p e r i m e n t a l l y
d e t e r m i n e d c o m p o s i t i o n of 64, 25, a n d 11 in wt% [46].
3.18. A r e s i d u e f r o m a N o r t h Sea crude has the following ex-
p e r i m e n t a l l y d e t e r m i n e d characteristics: 1)99(210) = 14.77
cSt., SG ----0.9217. F o r this fraction e s t i m a t e the follow-
ing p r o p e r t i e s a n d c o m p a r e with the e x p e r i m e n t a l val-
ues w h e r e they are available [46]:
a. K i n e m a t i c viscosity at 38~ (100~
b. M o l e c u l a r weight a n d average boiling p o i n t

Vol% IBP 5 10 20 30 40 50 60 70 80 90 95
TBP, ~ 420 451 470 495 514 ... 544 ... 580 ... 604 621
d20 ... 0.828 0.835 0.843 0.847 ... 0.851 ... 0.855 ... 0.856 0.859

3.12. A jet n a p h t h a has ASTM 50% t e m p e r a t u r e of 321 ~ a n d

specific gravity of 0.8046 [i]. The PNA c o m p o s i t i o n of
this fraction is 19, 70, a n d 11%, respectively. E s t i m a t e M,
n20, dzo, To, Pc, a n d Vc f r o m Tb a n d SG using the following
methods:
a. R i a z i - D a u b e r t (1980) m e t h o d s [38].
b. API-TDB m e t h o d s [2].
c. Twu m e t h o d s for M, Tc, Pc, a n d Vc.
d. Kesler-Lee m e t h o d s for M, Tc, a n d Pc.
e. S i m - D a u b e r t ( c o m p u t e r i z e d W i n n n o m o g r a p h )
m e t h o d for M, Tc, a n d Pc.
f. P s e u d o c o m p o n e n t m e t h o d .
c. Density a n d refractive index at 20~
d. P o u r p o i n t (experimental value is 39~
e. Sulfur c o n t e n t (experimental value is 0.63 wt%)
f. C o n r a d s o n c a r b o n r e s i d u e (experimental value is 4.6
wt%)
3.19. A c r u d e oil from N o r t h w e s t Australian field has total ni-
t r o g e n c o n t e n t of 310 p p m . One of the p r o d u c t s of a t m o -
spheric distillation c o l u m n for this c r u d e has true boil-
ing p o i n t range of 190-230~ a n d the API gravity of 45.5.
F o r this p r o d u c t d e t e r m i n e the following p r o p e r t i e s a n d
148 CHARACTERIZATION AND P R O P E R TI E S OF PETROLEUM FRACTIONS

TABLE 3.32--Basic parameters and sulfur content of some undefined petroleum products (Problem 3.15). a
(i) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Tb, K n2o SG, SG, M, M, d20, d20, rr~ RI Sulfur %, Sulfur %,
Fraction exp. exp. exp. calc. exp. calc. exp. calc. calc. calc. exp. calc.
Kuwaiti kerosene 468 1 . 4 4 1 0.791 . . . . . . 0.01
Kuwaiti diesel oil 583 1.480 0.860 1.3
US jet naphtha 434 1.444 0.805 144 01861 1.0
US high boiling naphtha 435 1.426 0.762 142.4 0.759 0.0
US kerosene 480 1.444 0.808 162.3 0.804 0.0
US fuel off 559 1.478 0.862 227.5 0.858 1.3
aExperimental data for the first two Kuwaiti fractions are taken from Riazi and Roomi [62] and for the US fractions are taken from Lenoir
and Hipkin data set [ 1].

c o m p a r e with the e x p e r i m e n t a l values as given b e l o w of 33 [46]. R e p o r t e d W a t s o n K factor is 11.72. Calculate

[46].
a. W h a t is the type of this p r o d u c t ?
b. M o l e c u l a r weight (experimental value is 160.1)
c. K i n e m a t i c viscosities at 20 a n d 40~ (experimental
values are 1.892 a n d 1.28)
d. F l a s h p o i n t (experimental value is 69~
e. PNA c o m p o s i t i o n (the e x p e r i m e n t a l PNA c o m p o s i -
tion from G C - M S analysis in vol% are 46.7, 40.5, a n d
12.8, respectively)
f. S m o k e p o i n t (experimental value is 26 m m )
g. Aniline p o i n t (experimental value is 64.9~
h. P o u r p o i n t (experimental value is - 4 5 ~
i. Freezing p o i n t (experimental value is - 4 4 ~
j. H y d r o g e n content (experimental value is 14.11 wt%)
3.20. An a t m o s p h e r i c residue p r o d u c e d f r o m the s a m e crude
of P r o b l e m 3.19 has API gravity of 25 a n d UOP K factor
of 12.0. Predict the following p r o p e r t i e s a n d c o m p a r e
with the e x p e r i m e n t a l values [46].
a. M o l e c u l a r weight (experimental value is 399.8)
b. Total nitrogen c o n t e n t (experimental value is 0.21
wt%)
c. K i n e m a t i c viscosity at 100~ (experimental value is
8.082)
d. K i n e m a t i c viscosity at 70~ (experimental value is
17.89)
e. Sulfur c o n t e n t (experimental value is 0.17 wt%)
f. C o n r a d s o n c a r b o n r e s i d u e (experimental value is 2.2
wt%)
g. C a r b o n content (experimental value is 86.7 wt%)
h. H y d r o g e n content (experimental value is 13 wt%)
i. Aniline p o i n t (experimental value is 95.2~
j. P o u r p o i n t (experimental value is 39~
3.21. A gas off p r o d u c e d from a c r u d e from S o r o o s h field
(Iran) has boiling r a n g e of 520-650~ a n d the API gravity
the following p r o p e r t i e s a n d c o m p a r e w i t h experimen-
tal values.
a. Average boiling p o i n t from Kw a n d c o m p a r e with m i d
boiling p o i n t
b. Cetane index ( r e p o r t e d value is 50.5)
c. Aniline p o i n t ( r e p o r t e d value is 152.9~
3.22. A heavy n a p h t h a s a m p l e f r o m Australian c r u d e oil
has the boiling range of 140-190~ specific gravity of
0.7736 a n d m o l e c u l a r weight of 131.4. The experimen-
tally d e t e r m i n e d c o m p o s i t i o n for n-paraffins, isoparaf-
fins, n a p h t h e n e s , a n d a r o m a t i c s are 29.97, 20.31, 38.72,
a n d 13%, respectively. F o r this s a m p l e the e x p e r i m e n t a l
values of RON a n d M O N are 26 a n d 28, respectively as
r e p o r t e d in Ref. [46], p. 359. E s t i m a t e the RON from the
p s e u d o c o m p o n e n t m e t h o d ( Eqs. 3.138 a n d 3.139) a n d
the Nelson m e t h o d s (Fig. 3.33 a n d 3.34). Also calculate
the MON f r o m Jenkins m e t h o d (Eq. 3.140). F o r each
case calculate the e r r o r a n d c o m m e n t o n y o u r results.
3.23. A light n a p h t h a f r o m Abu Dhabi field (UAE) has boil-
ing range of C5-80~ the API gravity of 83.1 a n d Kw of
12.73 [46]. E s t i m a t e the following octane n u m b e r s a n d
c o m p a r e w i t h the e x p e r i m e n t a l values.
a. Clear RON from two different m e t h o d s (experimental
value is 65)
b. RON + 1.5 m L of TEL /U.S. Gallon (experimental
value is 74.5)
c. Clear MON (experimental value is 61)
d. H o w m u c h MTBE s h o u l d be a d d e d to this n a p h t h a to
increase the RON f r o m 65 to 75.
3.24. A p e t r o l e u m fraction p r o d u c e d from a Venezuelan c r u d e
has ASTM D 86 distillation curve as:
vol% distilled 10 30 50 70 90
ASTM D 86 temperature, ~ 504 509 515 523 534

TABLE 3.33--Estimation of composition of petroleum fractions (Problem 3.16). a

Exp. composition Estimated composition
No, Fraction Tb, K exp. SG exp. n20 M m Xp XN XA Xp XN XA
1 China Heavy Naphtha 444.1 0.791 48.9 30.9 20.2
2 Malaysia Light Naphtha 326.6 0.666 83.0 17.0 0.0
3 Indonesia Heavy Naphtha 405.2 0.738 62.0 30.0 8.0
4 Venezuela Kerosene 463.6 0.806 39.8 41.1 19.0
5 Heavy Iranian Gasoline 323.5 0.647 93.5 5.7 0.8
6 Qatar Gasoline 309.4 0.649 95.0 3.9 1.1
7 Sharjah Gasoline 337.7 0.693 78.4 14.4 7.2
8 American Gasoline 317.2 0.653 92.0 7.3 0.7
9 Libya Kerosene 465.5 0.794 51.2 34.7 14.1
10 U.K. North Sea Kerosene 464.9 0.798 42.5 36.4 21.1
11 U.K. North Sea Gas Oil 574.7 0.855 34.3 39.8 25.9
12 Mexico Naphtha 324.7 0.677 81.9 13.9 4.2
aExperimental data on Tb, S, and the composition are taken from the Oil & Gas Journal Data Book [46].
 3. CHARACTERIZATION OF PETROLEUM FRACTIONS
The specific gravity (at 60/60~ ) of fraction is 0.8597
[46]. Predict the following p r o p e r t i e s for the p r o d u c t
a n d c o m p a r e with r e p o r t e d values.
a. K i n e m a t i c viscosities at 100, 140, a n d 210~ (experi-
m e n t a l values are 3.26, 2.04, a n d 1.12 cSt.)
b. Molecular weight f r o m viscosity a n d c o m p a r e with
m o l e c u l a r weight e s t i m a t e d f r o m the boiling point.
c. Boiling p o i n t a n d specific gravity from experimen-
tal values of k i n e m a t i c viscosities at 100 a n d 210~
(Part a) a n d c o m p a r e with actual values.
d. Aniline p o i n t (experimental value is 143.5~
e. Cetane n u m b e r (experimental value is 43.2)
f. Freezing p o i n t (experimental value is 21 ~
g. F l a s h p o i n t (experimental value is 230~
h. C a r b o n - t o - h y d r o g e n weight ratio (experimental value
is 6.69)
i. A r o m a t i c c o n t e n t f r o m e x p e r i m e n t a l viscosity at
100~ (experimental value is 34.9)
j. A r o m a t i c content from e x p e r i m e n t a l aniline p o i n t
(experimental value is 34.9)
k. Refractive index at 75~ (experimental value is
1.4759)
3.25. A v a c u u m r e s i d u e has k i n e m a t i c viscosity of 4.5 mm2/s
at 100~ a n d specific gravity of 0.854. E s t i m a t e viscosity
index (VI) of this fraction a n d c o m p a r e with r e p o r t e d
value of 119 [24].
3.26. A kerosene s a m p l e h a s boiling r a n g e of 180-225~ a n d
specific gravity of 0.793. This p r o d u c t has a r o m a t i c
c o n t e n t of 20.5%. Predict s m o k e p o i n t a n d freezing
p o i n t of this p r o d u c t a n d c o m p a r e with the experi-
m e n t a l values of 19 m m a n d - 5 0 ~ [24]. H o w m u c h
2 - m e t h y l n o n a n e (C10Hzz) with freezing p o i n t of - 7 4 ~
s h o u l d be a d d e d to this kerosene to r e d u c e the freezing
p o i n t to -60~
REFERENCES
[1] Lenoir, J. M. and Hipkin, H. G., "Measured Enthalpies of Eight
Hydrocarbon Fractions," Journal of Chemical and Engineering
Data, Vol. 18, No. 2, 1973, pp. 195-202.
[2] Daubert, T. E. and Danner, R. E, Eds., API Technical Data
Book Petroleum Refining, 6th ed., American Petroleum
Institute (API), Washington, DC, 1997.
[3] Curvers, J. and van den Engel, E, "Gas Chromatographic
Method for Simulated Distillation up to a Boiling Point of
750~ Using Temperature-Programmed Injection and High
Temperature Fused Silica Wide-Bore Columns," J. High
Resolution Chromatography, Vol. 12, 1989, pp. 16-22.
[4] ASTM, Annual Book of Standards, ASTM International, West
Conshohocken, PA, 2002.
[5] Denis, J., Briant, J. and Hipeaux, J. C., Lubricant Properties
Analysis and Testing, Translated to English by G. Dobson,
Editions Technip, Paris, 1997.
[6] Cryette, W. R., Model 5009 Wide Range Cryoscope--Instruction
Manual Precision Systems Inc., Natick, MA, 1994.
[7] Speight, J. G., The Chemistry and Technology of Petroleum, 3rd
ed., Marcel Dekker, New York, 1998.
[8] Firoozabadi, A., Thermodynamics of Hydrocarbon Reservoirs,
Mc-Graw Hill, New York, 1999.
[9] Smith, R. L. and Watson, K. M., "Boiling Points and Critical
Properties of Hydrocarbon Mixtures," Industrial and
Engineering Chemistry, Vol. 29, 1937, pp. 1408-1414.
149
[10] Zhou, E, "Correlation of the Average Boiling Points of
Petroleum Fractions with Pseudocritical Constants,"
International Chemistry Engineering, Vol. 24, 1984, pp. 731-742.
[11] Daubert, T. E., "Property Predictions," Hydrocarbon Processing,
March, 1980, pp. 107-112.
[12] Van Winkle, M., "Physical Properties of Petroleum Fraction,"
Petroleum Refiner, Vol. 34, No. 6, 1955, pp. 136-138.
[13] Nelson, W. L. and Hansburg, M., "Relation of ASTM and True
Boiling Point Distillations," Oil & Gas Journal, Vol. 38, No. 12,
1939, pp. 45-47.
[14] Edmister, W. C. and Pllock, D. H., "Phase Relations for
Petroleum Fractions," Chemical Engineering Progress, Vol. 44,
1948, pp. 905-926.
[15] Edmister, W. C. and Okamoto, K. K., "Applied Hydrocarbon
Thermodynamics, Part 12: Equilibrium Flash Vaporization
Correlations for Petroleum Fractions," Petroleum Refiner,
Vol. 38, No. 8, 1959, pp. 117-132.
[16] Edmister, W. C. and Okamoto, K. K., Applied Hydrocarbon
Thermodynamics, Gulf Publishing, Houston, TX, 1961,
pp. 116-133.
[17] Edmister, W. C. and Okamoto, K. K., "Applied Hydrocarbon
Thermodynamics, Part 13: Equilibrium Flash Vaporization
Correlations for Heavy Oils Under Subatmospheric Pressures,"
Petroleum Refiner, Vol. 38, No. 9, 1959, pp. 271-288.
[18] Okamoto, K. K. and van Winkle, M., "Equilibrium Flash
Vaporization of Petroleum Fractions," Industrial and
Engineering Chemistry, Vol. 45, 1953, pp. 429-439.
[19] House, H. G., Braun, W. G., Thompson, W. T. and Fenske,
M. R , "Documentation of the Basis for Selection of the
Contents of Chapter 3, ASTM, TBP, and EFV Relationships for
Petroleum Fractions in Technical Data Book-Petroleum
Refining," Documentation Report No. 3-66, Pennsylvania State
University, 1966.
[20] Arnold, V. E., "Microcomputer Program Converts TBP, ASTM,
EFV Distillation Curves," Oil & Gas Journal, Vol. 83, No. 6,
1985, pp. 55-62.
[21] Ford, D. C., Miller, W. H., Thren, R. C. and Wertzler, R.,
Correlation of ASTM Method D 2887-73 Boiling Range
Distribution Data with ASTM Method D 86-67 Distillation Data,
STP 577, ASTM International, West Conshohocken, PA, 1973,
p. 29.
[22] Riazi, M. R. and Daubert, T. E., "Analytical Correlations
Interconvert Distillation Curve Types," Oil & Gas Journal,
Vol. 84, 1986, August 25, pp. 50-57.
[23] Daubert, T. E.,"Petroleum Fraction Distillation
Interconversion," Hydrocarbon Processing, Vol. 73, No. 9, 1994,
pp. 75-78.
[24] Wauquier, J.-E, "Petroleum Refining, Vol. 1: Crude Oil,
Petroleum Products, Process Ftowsheets," Editions Technip,
Paris, 1995.
[25] HYSYS, "Reference Volume 1., Version 1.1," HYSYS Reference
Manual for Computer Software, HYSYS Conceptual Design,
Hyprotech Ltd., Calgary, Alberta, Canada, 1996.
[26] Daubert, T. E., Riazi, M. R. and Danner, R. E, "Documentation
of the Basis for Selection of the Contents of Chapters 2 and 3 in
the API Technical Data Book-Petroleum Refining,"
Documentation Report No. 2,3-86, Pennsylvania State
University, 1986.
[27] Maxwell, J. B. and Bonnel, L. S., Vapor Pressure Charts for
Petroleum Engineers, Esso Research and Engineering Company,
Linden, NJ, 1955.
[28] Myers, H. S. and Fenske, M. R., "Measurement and Correlation
of Vapor Pressure Data for High-Boiling Hydrocarbons,"
Industrial and Engineering Chemistry, Vol. 47, 1955,
pp. 1652-1658.
[29] Riazi, M. R., Prediction of Thermophysical Properties of
Petroleum Fractions, Doctoral Dissertation, Department of
150 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
Chemical Engineering, Pennsylvania State University,
University Park, PA, 1979.
[30] Van Nes, K. and Van Western, H. A., Aspects of the Constitution
of Mineral Oils, Elsevier, New York, 1951.
[31] Riazi, M. R., "Distribution Model for Properties of
Hydrocarbon-Plus Fractions," Industrial and Engineering
Chemistry Research, Vol. 28, 1989, pp. 1731-1735.
[32] Riazi, M. R. and Daubert, T. E., "Improved Characterization of
Wide Boiling Range Undefined Petroleum Fractions," Industrial
and Engineering Chemistry Research, Vol. 26, 1987, pp. 629-632.
[33] Huang, E K. and Daubert, T. E., "Prediction of the Enthalpy of
Petroleum Fractions: Pseudocompound Method," Industrial
and Engineering Chemistry, Process Design and Development,
Vol. 13, No. 4, 1974, pp. 359-362.
[34] Katinas, T. G., Prediction of the Viscosities of Petroleum
Fractions, M. Sc. Thesis, Department of Chemical Engineering,
Pennsylvania State University, University Park, PA,
1977.
[35] Miqueu, C., Satherley, J., Mendiboure, B., Lachaise, J. and
Graciaa, A., "The Effect of P/N/A Distribution on the Parachors
of Petroleum Fractions," Fluid Phase Equilibria, Vol. 180,
No.1/2, 2001, pp. 327-344.
[36] Riazi, M. R. and Daubert, T. E., "Prediction of Molecular Type
Analysis of Petroleum Fractions and Coal Liquids," Industrial
and Engineering Chemistry, Process Design and Development,
Vol. 25, No. 4, 1986, pp. 1009-1015.
[37] Riazi, M. R. and A1-Sahhaf, T., "Physical Properties of n-Alkanes
and n-Alkyl Hydrocarbons: Application to Petroleum Mixtures,"
Industrial and Engineering Chemistry Research, Vol. 34, 1995,
pp. 4145-4148.
[38] Riazi, M. R. and Daubert, T. E., "Simplify Property Predictions,"
Hydrocarbon Processing, Vol. 59, No. 3, 1980, pp. 115-116.
[39] Lee, B. I. and Kesler, M. G., "A Generalized Thermodynamic
Correlation Based on Three- Parameter Corresponding States,"
American Institute of Chemical Engineers Journal, Vol. 21, 1975,
pp. 510-527.
[40] Kesler, M. G. and Lee, B. I., "Improve Prediction of Enthalpy of
Fractions," Hydrocarbon Processing, Vol. 55, No. 3, March 1976,
pp. 153-158.
[41] Winn, E W., "Physical Properties by Nomogram," Petroleum
Refiners, Vol. 36, No. 21, 1957, pp. 157-159.
[42] Twu, C. H., "An Internally Consistent Correlation for Predicting
the Critical Properties and Molecular Weights of Petroleum
and Coal-Tar Liquids," Fluid Phase Equilbria, Vol. 16, 1984,
pp. 137-150.
[43] Goossens, A. G., "Prediction of Molecular Weight of Petroleum
Fractions," Industrial and Engineering Chemistry Research, Vol.
35, 1996, pp. 985-988.
[44] Riazi, M. R. and Daubert, T. E., "Characterization Parameters
for Petroleum Fractions," Industrial and Engineering
Chemistry Research, Vol. 26, 1987, pp. 755-759. (Corrections,
p. 1268.)
[45] Riazi, M. R. and Roomi, Y., "Use of the Refractive Index in the
Estimation of Thermophysical Properties of Hydrocarbons and
Their Mixtures," Industrial and Engineering Chemistry
Research, Vol. 40, No. 8, 2001, pp. 1975-1984.
[46] Oil and Gas Journal Data Book, 2000 Edition, PennWell, Tulsa,
OK, 2000.
[47] Riazi, M. R. and Daubert, T. E., "Prediction of the Composition
of Petroleum Fractions," Industrial and Engineering Chemistry,
Process Design and Development, Vol. 19, No. 2, 1980,
pp. 289-294.
[48] Ahmed, T., Hydrocarbon Phase Behavior, Gulf Publishing,
Houston, TX, 1989.
[49] Hill, J. B. and Coats, H. B., "The Viscosity Gravity Constant of
Petroleum Lubricating Oils," Industrial and Engineering
Chemistry, Vol. 20, 1928, pp. 641-644.
[50] Watson, K. M. and Nelson, E. F., "Improved Methods for
Approximating Critical and Thermal Properties of Petroleum
Fractions," Industrial and Engineering Chemistry, Vol. 25, 1933,
pp. 880-887.
[51] Watson, K. M., Nelson, E. E, and Murphy, G. B.,
"Characterization of Petroleum Fractions," Industrial and
Engineering Chemistry, Vol. 27, 1935, pp. 1460-1464.
[52] Kurtz, Jr., S. S. and Ward, A. L., "The Refractivity Intercept and
the Specific Refraction Equation of Newton, I: Development of
the Refractivity Intercept and Composition with Specific
Refraction Equations," Journal of Franklin Institute, Vol. 222,
1936, pp. 563-592.
[53] Huang, P. K., Characterization and Thermodynamic Correlations
for Undefined Hydrocarbon Mixtures, Ph.D. Dissertation,
Pennsylvania State University, University Park, PA, 1977.
[54] Fryback, M. G., "Synthetic Fuels: Promises and Problems,"
Chemical Engineering Progress, Vol. 77, No. 5, 1981,
pp. 39-43.
[55] Hersh, R. E., Fenske, M. R., Booser, E. R., and Koch, E. E,
"Ring Analysis of Hydrocarbon Mixtures," Journal of Institute
of Petroleum, Vol. 36, No. 322, 1950, pp. 624-668.
[56] API Research Project 42: Properties of Hydrocarbons of High
Molecular Weight, American Petroleum Institute, New York,
1966.
[57] Goodger, E. and Vere, R., Aviation Fuels Technology, MacMillan,
London, 1985.
[58] Goossens, A. G., "Prediction of the Hydrogen Content of
Petroleum Fractions," Industrial and Engineering Chemistry
Research, Vol. 36, 1997, pp. 2500-2504.
[59] Jenkins, G. I. and Walsh, R. E, "Quick Measure of Jet Fuel
Properties," Hydrocarbon Processing, Vol. 47, No. 5, 1968,
pp. 161-164.
[60] Fein, R. S., Wilson, H. I., and Sherman, J., "Net Heat of
Combustion of Petroleum Fractions," Industrial and
Engineering Chemistry, Vol. 45, No. 3, 1953, pp. 610-614.
[61] Baird, C. T., Crude Oil Yields and Product Properties, Ch. De la
Haute-Belotte 6, Cud Thomas Baird IV, 1222 Vezenaz, Geneva,
Switzerland, June, 1981.
[62] Riazi, M. R., Nasimi, N., and Roomi, Y., "Estimating Sulfur
Content of Petroleum Products and Crude Oils," Industrial and
Engineering Chemistry Research, Vol. 38, No. 11, 1999,
pp. 4507-4512.
[63] Gary, J. H. and Handwerk, G. E., Petroleum Refining, Technology
and Economic, 3rd ed., Marcel Dekker, New York, 1994.
[64] Hebert, J., "Court Upholds Tougher Diesel Rules," Associated
Press Writer, Yahoo News Website, May 3, 2002.
[65] Ball, S. J., Whisman, M. L., and Wenger, W. J., "Nitrogen
Content of Crude Petroleums," Industrial and Engineering
Chemistry, Vol. 43, No. 11, 1951, pp. 2577-2581.
[66] Nelson, W. L., Petroleum Refinery Engineering, 4th ed.,
McGraw-Hill, New York, 1969.
[67] Riazi, M. R., Albahri, T. A., and Alqattan, A., "Prediction of Reid
Vapor Pressure of Petroleum Fuels," Petroleum Science and
Technology, Vol. 23, No. 1, 2005, pp. 1-12.
[68] Albahri, T., Riazi, M. R., and Alqattan, A., "Analysis of Quality
of the Petroleum Fuels," Journal of Energy and Fuels, Vol. 17,
No. 3, 2003, pp. 689-693.
[69] Hatzioznnidis, I., Voutsas, E. C., Lois, E., and Tassios, D. P.,
"Measurement and Prediction of Reid Vapor Pressure of
Gasoline in the Presence of Additives," Journal of Chemical and
Engineering Data, Vol. 43, No. 3, 1998, pp. 386-392.
[70] Butler, R. M., Cooke, G. M., Lukk, G. G., and Jameson. B. G.,
"Prediction of Flash Points of Middle Distillates," Industrial
and Engineering Chemistry, Vol. 48, No. 4, 1956, pp. 808- 812.
[71] Walsh, R. E and Mortimer, J. V., "New Way to Test Product
Quality," Hydrocarbon Processing, Vol. 50, No. 9, 1971,
pp. 153-158.
 MNL50-EB/Jan. 2005

Appendix
ASTM DEFINITIONS OF TERMS
ASTM DICTIONARY OF SCIENCE AND TECHNOLOGY1 defines vari-
o u s engineering terms in standard terminology. ASTM pro-
vides several definitions for most properties by its different
committees. The closest definitions to the properties used in
the book are given below. The identifier provided includes the
standard designation in which the term appears followed by
the committee having jurisdiction of that standard. For ex-
ample, D02 represents the ASTM Committee on Petroleum
Products and Lubricants.
Additive---Any substance added in small quantities to an-
other substance, usually to improve properties; sometimes
called a modifier. D 16, D01
Aniline p o i n t m T h e minimum equilibrium solution temper-
ature for equal volumes of aniline (aminobenzene) and
sample. D 4175, D02
API g r a v i t y ~ A n arbitrary scale developed by the American
Petroleum Institute and frequently used in reference
to petroleum insulating oil. The relationship between API
gravity and specific gravity 60/60~ is defined by the fol-
lowing: Degree API gravity at 60~ = 141.5/(SG 60/60~ -
131.5. [Note: For definition see Eq. (2.4) in this book.]
D 2864, D27
Ash--Residue after the combustion of a substance under
specified conditions. D 2652, D28
Assay~Analysis of a mixture to determine the presence or
concentration of a particular component. F 1494, F23
Autoignition--The ignition of material caused by the appli-
cation of pressure, heat, or radiation, rather than by an
external ignition source, such as a spark, flame, or incan-
descent surface. D 4175, D02
Autoignition t e m p e r a t u r e - - T h e minimum temperature at
which autoignition occurs. D 4175, D02
Average (for a series of observations)~The total divided by
the number of observations. D123, D13
Bar--Unit of pressure; 14.5 lb/in 2, 1.020 kg/cm 2, 0.987 atm,
0.1 MPa. D 6161, D I 9
Bitumen--A class of black or dark-colored (solid, semisolid,
or viscous) cementitious substances, natural or manufac-
tured, composed principally of high-molecular-weight hy-
drocarbons, of which asphalts, tars, pitches, and asphaltites
are typical. D 8, D04
Boiling p o i n t - - T h e temperature at which the vapor pressure
of an engine coolant reaches atmospheric pressure under
equilibrium boiling conditions. [Note: This definition is ap-
plicable to all types of liquids.] D 4725, D I S
Boiling p r e s s u r e - - A t a specified temperature, the pressure
at which a liquid and its vapor are in equilibrium.
E 7, E04
ASTM Dictionary of Engineering Science and Technology, 9th ed.,
ASTM International, West Conshohocken, PA, 2000.
397
BTU--One British thermal unit is the amount of heat re-
quired to raise 1 lb of water I~ E 1705, E48
Carbon blackmA material consisting essentially of elemental
carbon in the form of near-spherical colloidal particles and
coalesced particle aggregates of colloidal size, obtained by
partial combustion or thermal decomposition of hydrocar-
bons. D 1566, D l l
Carbon residue---The residue formed by evaporation and
thermal degradation of a carbon-containing material.
D 4175, D02
Catalyst--A substance whose presence initiates or changes
the rate of a chemical reaction, but does not itself enter
into the reaction. C 904, C03
Cetane number (cn)----A measure of the ignition perfor-
mance of a diesel fuel obtained by comparing it to reference
fuels in a standardized engine test. D 4175, D02
Chemical p o t e n t i a l (/~i or r partial molar free energy
of component i, that is, the change in the free energy of a
solution upon adding 1 tool of component i to an infinite
amount of solution of given composition, (SG/~n4)r.v,.~ =
Gi =/zi, where G -- Gibbs free energy and r~ = number of
moles of the ith component. E 7, E04
Cloud p o i n t - - T h e temperature at which a defined liquid mix-
ture, under controlled cooling, produces perceptible haze
or cloudiness due to the formation of fine particles of an
incompatible material. D 6440, D01
Coal--A brown to black combustible sedimentary rock (in
the geological sense) composed principally of consolidated
and chemically altered plant remains. D 121, D05
CokemA carbonaceous solid produced from coal, petroleum,
or other materials by thermal decomposition with passage
through a plastic state. C 709, D02
Combustion~A chemical process of oxidation that occurs at
a rate fast enough to produce heat and usually light either
as glow or flames. D 123, D I 3
Compressed natural gas (CNG)~Natural gas that is typi-
cally pressurized to 3600 psi. CNG is primarily used as a
vehicular fuel. D 4150, D03
Concentration--Quantity of substance in a unit quantity of
sample. E 1605, E06
Critical p o i n t - - I n a phase diagram, that specific value
of composition, temperature, pressure, or combinations
thereof at which the phases of a heterogeneous equilibrium
become identical. E 7, E04
Critical pressure---Pressure at the critical point.
E 1142, E37
Critical temperature----(1) Temperature above which the
vapor phase cannot be condensed to liquid by an increase
in pressure. E 7, E04
(2)Temperature at the critical point. E 1142, E37
D e g r a d a t i o n ~ D a m a g e by weakening or loss of some prop-
erty, quality, or capability. E 1749, E 06

Copyright 9 2005 by ASTM International www.astm.org

398 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S
Degree Celsius (~ unit of temperature in the In-
ternational System of Units (SI). E 344, E20
Density--The mass per unit volume of a substrate at a spec-
ified temperature and pressure; usually expressed in g/mE
kg/L, g/cm 3, g/L, kg/m 3, or lb/gal. D 16, D01
Deposition--The chemical, mechanical, or biological pro-
cesses through which sediments accumulate in a resting
place. D 4410, D19
Dew point--The temperature at any given pressure at which
liquid initially condenses from a gas or vapor. It is specifi-
cally applied to the temperature at which water vapor starts
to condense from a gas mixture (water dew point) or at
which hydrocarbons start to condense (hydrocarbon dew
point). D 4150, D03
Diffusion--(1) Spreading of a constituent in a gas, liquid, or
solid tending to make the composition of all parts uniform.
(2) The spontaneous movement of atoms or molecules to
new sites within a material. B 374, B08
Distillation--The act of vaporizing and condensing a liq-
uid in sequential steps to effect separation from a liquid
mixture. E 1705, E 48
Distillation t e m p e r a t u r e (in a c o l u n m distillation)---The
temperature of the saturated vapor measured just above
the top of the fractionating column. D 4175, D02
E n d o t h e r m i e r e a c t i o n - - A chemical reaction in which heat
is absorbed. C 1145, C 28
E n t h a l p y m A thermodynamic function defined by the equa-
tion H = U + PV, where H is the enthalpy, U is the internal
energy, P is the pressure, and V the volume of the system.
[Note: Also see Eq. (6.1) of this book.] E 1142, E37
E q u i l i b r i u m ~ A state of dynamic balance between the op-
posing actions, reactions, or velocities of a reversible
process. E 7, E04
Evaporation--Process where a liquid (water) passes from a
liquid to a gaseous state. D 6161, D19
Fire p o i n t - - T h e lowest temperature at which a liquid or solid
specimen will sustain burning for 5 s. D 4175, D02
Flammable liquid--A liquid having a flash point below
37.8~ (100~ and having a vapor pressure not exceed-
ing 40 psi (absolute) at 37.8~ and known as a Class I liquid.
E 772, E44
Flash p o i n t - - T h e lowest temperature of a specimen cor-
rected to a pressure of 760 m m H g (101.3 kPa), at which
application of an ignition source causes any vapor from
the specimen to ignite under specified conditions of test.
D 1711, D09
FluiditymThe reciprocal of viscosity. D 1695, D01
Freezing point--The temperature at which the liquid and
solid states of a substance are in equilibrium at a given
pressure (usually atmospheric). For pure substances it is
identical with the melting point of the solid form.
D 4790, D16
Gas--One of the states of matter, having neither independent
shape nor volume and tending to expand indefinitely.
D 1356, D22
Gasification--Any chemical or heat process used to convert
a feedstock to a gaseous fuel. E 1126, E 48
Gasoline---A volatile mixture of liquid hydrocarbons, nor-
mally containing small amounts of additives, suitable for
use as a fuel in spark-ignition internal combustion engines.
D 4175, D02
Gibbs free energy--The maximum useful work that can be
obtained from a chemical system without net change in
temperature or pressure, AF = AH - TAS. [Note: For def-
inition see Eq. (6.6) in this book; the author has used G for
Gibbs free energy.] E 7, E04
GrainmUnit of weight; 0.648 g, 0.000143 lb. D 6161, D19
Gross calorific value (synonym: higher heating value,
HI-IV)raThe energy released by combustion of a unit quan-
tity of refuse-derived fuel at constant volume or constant
pressure in a suitable calorimeter under specified condi-
tions such that all water in the products is in liquid form.
This the measure of calorific value is predominately used
in the United States. E 856, D34
Heat capacity--The quantity of heat required to raise a sys-
tem 1~in temperature either at constant volume or constant
pressure. D 5681, D34
Heat flux (q)~The heat flow rate through a surface of unit
area perpendicular to the direction of heat flow (q in SI
units: W/m 2; q in inch-pound units: Btu/h/ft 2 = Btu/h 9ft2)
C 168, C16
Henry's law--The principle that the mass of a gas dissolved
in a liquid is proportional to the pressure of the gas above
the liquid. D 4175, D02
Higher heating value (HHV)--A synonym for gross calorific
value. D 5681, D34
Inert components---Those elements or components of nat-
ural gas (fuel gas) that do not contribute to the heating
value. D 4150, D03
Inhibitor--A substance added to a material to retard or pre-
vent deterioration. D 4790, D16
Initial boiling p o i n t - - T h e temperature observed immedi-
ately after the first drop of distillate falls into the receiving
cylinder during a distillation test. D 4790, D 16
Interface--A boundary between two phases with different
chemical or physical properties. E 673, E 42
Interracial tension (IF]F)---The force existing in a liquid-
liquid phase interface that tends to diminish the area of
the interface. This force, which is analogous to the surface
tension of liquid-vapor interfaces, acts at each point on the
interface in the plane tangent at that point. D 459, D I 2
International System of Units, SI--A complete coherent
system of units whose base units are the meter, kilogram,
second, ampere, kelvin, mole, and candela. Other units are
derived as combinations of the base units or are supple-
mentary units. A 340, A06
Interphase---The region between two distinct phases over
which there is a variation of a property. E 673, E42
ISO--Abbreviation for International Organization for Stan-
dards: An organization that develops and publishes inter-
national standards for a variety of technical applications,
including data processing and communications.
E 1457, F05
Jet fuel--Any liquid suitable for the generation of power by
combustion in aircraft gas turbine engines. D 4175, D02
Joule (J)--The unit of energy in the SI system of units. One
joule is 1 W..-. A 340, A06
Kelvin (K)--The unit of thermodynamic temperature; the SI
unit of temperature for which an interval of 1 kelvin (K)
equals exactly an interval of 1~ and for which a level of
273.15 K equals exactly 0~ D 123, D13

Liquefied petroleum gas (LPG)--A mixture of normally
gaseous hydrocarbons, predominantly propane or butane
or both, that has been liquefied by compression or cooling,
or both, to facilitate storage, transport, and handling.
D 4175, D02
Liquid--A substance that has a definite volume but no defi-
nite form, except such given by its container. It has a viscos-
ity of 1 x 10-3 to 1 x 103 St (1 x 10-7 to 1 x 10-1 m 2 "S-1)
at 104~ (40~ or an equivalent viscosity at agreed upon
temperature. (This does not include powders and granular
materials.) Liquids are divided into two classes:
(1) Class A, low viscosity--A liquid having a viscosity of
1 x 10-3 to 25.00 St (1 x 10 -7 to 25.00 x 10-4 m 2 -s -I) at
104~ (40~ or an equivalent viscosity at agreed upon
temperature.
(2) Class B, high viscosity--A liquid having a viscosity of
25.01 to 1 x 103 St (25.01 x 10-4 to 1 x 10-1 m 2.s 1)
at 104~ (40~ or an equivalent viscosity at agreed upon
temperature.
D 16, D01
Lower heating value (LHV)mA synonym for net calorific
value. D 5681, D34
Lubricant--Any material interposed between two surfaces
that reduces the friction or wear between them.
D 4175, D02
MassmThe quantity of matter in a body (also see weight).
D 123, D13
Melting p o i n t - - I n a phase diagram, the temperature at
which the liquids and solids coincide at an invariant point.
E 7, E04
Micron (/~m, micrometer)--A metric unit of measurement
equivalent to 10-6 m, 10 4 cm.
1) 6161, D19
Molality~Moles (gram molecular weight) of solute per
1000 g of solvent. 1) 6161, 1)19
Molarity~Moles (gram molecular weight) of solute per liter
of total solution 1) 6161, 1)19
Molecular d i f f u s i o n ~ A process of spontaneous intermixing
of different substances, attributable to molecular motion,
and tending to produce uniformity of concentration.
D1356, D22
Mole fraction--The ratio of the number of molecules (or
moles) of a compound or element to the total number of
molecules (or moles) present. 1) 4023, 1)22
Naphtha, aromatic solvent--A concentrate of aromatic hy-
drocarbons including C8, C9, and C10 homologs.
D 4790, 1) 16
Napthenic oilmAn hydrocarbon process oil containing more
than 30%, by mass, of naphthenic hydrocarbons.
1) 1566, D l l
Natural gas---A naturally occurring mixture of hydrocarbon
and nonhydrocarbon gases found in porous geological for-
mations (reservoirs) beneath the earth's surface, often in
association with petroleum. The principal constituent of
natural gas is methane. 1) 4150, D03
Net calorific value (Net heat of combustion at constant
pressure)--The heat produced by combustion of unit
quantity of a solid or liquid fuel when burned, at constant
pressure of 1 atm (0.1 MPa), under the conditions such
APPENDIX 399
that all the water in the products remains in the form of
vapor. D 121, I)05
Net heat of combustion--The oxygen bomb (see Test
Method D 3286) value for the heat of combustion, corrected
for gaseous state of product water. E 176, E05
Octane number (for spark ignition engine fuel)mAny one
of several numerical indicators of resistance to knock ob-
tained by comparison with reference fuels in standardized
engine or vehicle tests. D 4175, D02
Oxygenate---An oxygen-containing ashless organic com-
pound, such as an alcohol or ether, which may be used as
a fuel or fuel supplement. D 4175, D02
Paraffinic oilmA petroleum oil (derived from paraffin crude
oil) whose paraffinic carbon type content is typically greater
than 60%. E 1519, E35
Partial pressure---The contribution of one component of a
system to the total pressure of its vapor at a specified tem-
perature and gross composition. E 7, E04
Porosity--The percentage of the total volume of a material
occupied by both open and closed pores. [Note: In this book
porosity represented by ~ (see Eq. 8.72) is the fraction of
total volume of a material occupied by open pores and is
not identical to this definition.] C 709, D02
Pour point--The lowest temperature at which a liquid can
be observed to flow under specified conditions.
1) 2864, 1)27
Precipitation--Separation of new phase from solid, liquid,
or gaseous solutions, usually with changing conditions or
temperature or pressure, or both. E 7, E04
Pressure--The internal force per unit area exerted by any
material. Since the pressure is directly dependent on the
temperature, the latter must be specified. 1) 3064, 1)10
Pressure, saturation--The pressure, for a pure substance
at any given temperature, at which vapor and liquid, or
vapor and solid, coexist in stable equilibrium. [Note: This
is the definition of vapor pressure used in this book.]
E 41, G03
QualitymCollection of features and characteristics of a prod-
uct, process, or service that confers its ability to satisfy
stated or implied needs. E 253, E18
Range---The region between the limits within which a quan-
tity is measured and is expressed by stating the lower and
upper range values. E 344, E20
Refractive index--The ratio of the velocity of light (of speci-
fied wavelength) in air to its velocity in the substance under
examination. This is relative refractive index of refraction.
If absolute refractive index (that is, referred to vacuum)
is desired, this value should be multiplied by the factor
1.00027, the absolute refractive index of air. [Note: In this
book absolute refractive index is used.] 1) 4175, 1)02
Saturation--The condition of coexistence in stable equilib-
rium of a vapor and a liquid or a vapor and solid phase of
the same substance at the same temperature. E 41, G03
S m o k e pointmThe maximum height of a smokeless flame of
fuel burned in a wick-fed lamp. 1) 4175, 1)02
SolidmA state of matter in which the relative motion of
molecules is restricted and in which molecules tend to
retain a definite fixed position relative to each other. A
solid may be said to have a definite shape and volume.
E 1547, E 15
400 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS

Solubility--The extent that one material will dissolve in an- Wax a p p e a r a n c e p o i n t - - T h e temperature at which wax or
other, generally expressed as mass percent, or as volume other solid substances first begin to separate from the liquid
percent, or parts per 100 parts of solvent by mass or vol- oil when it is cooled under prescribed conditions (refer to
ume. The temperature should be specified. D 3064, D10 D 3117, Test Method for Wax Appearance Point of Distillate
Solubility p a r a m e t e r (of liquids)--The square root of the Fuels). D 2864, D27
heat of vaporization minus work of vaporization (cohesive Weight ( s y n o n y m o u s with m a s s ) - - T h e mass of a body
energy density) per unit volume of liquid at 298 K. is a measure of its inertia, or resistance to change in motion.
D 4175, D02 E 867, E17
Solutes--Matter dissolved in a solvent. D 6161, D19
Specific gravity (deprecated term of liquids)--The ratio of
density of a substance to that of a reference substance such
as water (for solids and liquids) or hydrogen (for gases)
Greek Alphabet
under specified conditions. Also called relative density. Alpha
[Note: In this book the reference substance for definition fl Beta
of gas specific gravity is air]. D 4175, D02 F G a m m a (Uppercase)
Surface t e n s i o n - - P r o p e r t y that exists due to molecular F Gamma
forces in the surface film of all liquids and tends to pre- A Delta (Uppercase)
vent the liquid from spreading. B 374, B08 8 Delta
T e m p e r a t u r e - - T h e thermal state of matter as measured on e Epsilon
a definite scale. B 713, B01 ( Zeta
T h e r m a l conductivity (X)--Time rate of heat flow, under Eta
steady conditions, through unit area, per unit temperature | Theta (Uppercase)
gradient in the direction perpendicular to the area. 0 Theta
E 1142, E37 K Kappa (Uppercase)
T h e r m a l diffusivity--Ratio of thermal conductivity of a K Kappa
substance to the product of its density and specific heat A Lambda (Uppercase)
capacity. E l 142, E37 )~ Lamhda
Vapor--The gaseous phase of matter that normally exists in # Mu
a liquid or solid state. D 1356, D22 v Nu
Vapor pressure---The pressure exerted by the vapor of a liq- Xi
uid when in equilibrium with the liquid. D 4175, D02 H Pi (Uppercase)
Viscosity, absolute (~/)--The ratio of shear stress to shear zr Pi
rate. It is the property of internal resistance of a fluid that p Rho
opposes the relative motion of adjacent layers [Note: See N Sigma (Uppercase)
Eq. (8.1) in this book.] The unit most commonly used for Sigma
insulating fluids is centipoise. D 2864, D27 r Tau
Viscosity, k i n e m a t i c - - T h e quotient of the absolute (dy- v Upsilon
namic) viscosity divided by the density, O/p both at the q~ Phi (Uppercase)
same temperature. For insulating liquids, the unit most q~ Phi
commonly unit is the centistokes (100 cSt = 1 St). [Note: Phi
See Eq. (8.1) in this book.] D 2864, D27 x Chi
Viscosity, Saybolt Universal--The efflux time in seconds of Psi (Uppercase)
60 mL of sample flowing through a calibrated Saybolt Uni- Omega (Upper case)
versal orifice under specified conditions. D 2864, D27 co Omega
 MNL50-EB/Jan. 2005

Index
A ASTM D 92, 34, 133 ASTM method, 128
ASTM D 93, 34, 133-134, 144 molecular weight prediction, 56
Absolute density, 120 ASTM D 97, 135, 144 Atmospheric critical pressure, heavy
Academia, 17 ASTM D 129, 99 hydrocarbons, 51
Acentric factor, 11 ASTM D 189, 141,144 Autoignition temperature, definition, 34
aromatics, 52 ASTM D 240, 144 Avogadro number, 24
definition, 33 ASTM D 287, 93
estimation, 80-82, 115-116 ASTM D 323, 33, 144
pure hydrocarbons, prediction, 64-66 ASTM D 341, 70, 338 B
Activation energy, 346 ASTM D 357, 34, 139
Activity coefficients ASTM D 445, 100, 144, 338 Benedict-Webb-Rubin equation of state,
mixtures, 254-255 ASTM D 446, 338 modified, 214, 217-220
calculation, 257-261 ASTM D 524, 144, 141 Benzene, 4-5
Albahri et al. method, 137 ASTM D 611, 35, 137, 144 vapor pressure, 3 l 3
Alcohols, octane number, 139 ASTM D 613, 138 Binary interaction parameter, 209-210,
Alkanes ASTM D 908, 34, 139 269-270
boiling point, 58-59 ASTM D 976, 138 Binary systems, freezing-melting diagram,
critical compressibility factor, 64 ASTM D 1018, 99 285
critical temperature, 50 ASTM D 1160, 92, 100-101,106, 108, 110, Block and Bird correlation, 359
entropy of fusion, 262 114, 144 Boiling point, 11
liquid thermal conductivity, 343-344 ASTM D 1218 94, 144 n-alkanes, 58-59
surface tension, 361 ASTM D 1262 99 definition, 31
vapor pressure, 306 ASTM D 1266 99, 144 elevation, 282-284
n-Alkyl, critical pressure, 52 ASTM D 1298 93 heavy hydrocarbons, 50, 52
Alkylbenzene ASTM D 1319 144 hydrocarbon-plus fractions, 173
entropy of fusion, 262 ASTM D 1322 142 petroleum fractions, 88-93
vapor pressure, 307 ASTM D 1368 99 prediction, pure hydrocarbons, 58-59
Analytical instruments, 96-98 ASTM D 1500 144 reduced, 251
ANFOR M 15-023, 10 ASTM D 1548 99 sub- or superatmospheric pressures,
Aniline point, 11 ASTM D 1552 99 106-107
definition, 35 ASTM D 1747 95 true, 89
petroleum fractions, 137 ASTM D 2007 96 Boiling points
Antoine coefficients, 310 ASTM D 2267 10 average, 100-101
Antoine equation, 305-306 ASTM D 2270, 122-124 range, 88
API degree, 21 ASTM D 2386, 136, 144 Boiling range fractions, narrow versus
API gravity, 11 ASTM D 2500, 135, 144 wide, 112-119
crude oils, 156 ASTM D 2501, 36 Boltzman constant, 24
definition, 32 ASTM D 2502, 56 Boossens correlation, 57-58
petroleum fractions, 93 ASTM D 2503, 94 Bubble point, calculations, 370-371
prediction, pure hydrocarbons, ASTM D 2533, 133 Bubble point curve, 201
58-60 ASTM D 2549, 97 Bubble point pressure, 223, 367
API methods, 124-126 ASTM D 2700, 144 Bubble point temperature, 368
critical temperature and pressure, ASTM D 2717-95, 144 Bulk parameters, petroleum fractions, 114
prediction, 60 ASTM D 2759, 127 Butane, equilibrium ratios, 274-275
critical volume, prediciton, 63 ASTM D 2887, 12, 89-90, I00, 104-105,
molecular weight prediction, 56 110, 144
API RP 42, 37, 56 ASTM D 2890, 320-321 C
API Technical Data Book-Petroleum ASTM D 2892, 144, 154
Refining, 15 ASTM D 2983, 144 Capillary pressure, 357
Aromatics, 4-5 ASTM D 3178, 99 Carbon
Arrhenius-type equation, 346 ASTM D 3179, 99 prediction in petroleum fractions, 127
Asphalt, 10 ASTM D 3228, 99 see also SCN groups
Asphaltene, 373-378 ASTM D 3238, 121, 126 Carbon number range approach, petroleum
inhibitor, 377-378 ASTM D 3343, 99, 128, 130 fractions, 186
precipitation, 375, 377, 379 ASTM D 3431, 99 Carbon residue, petroleum fractions,
solid-liquid equilibrium, 385-388 ASTM D 3710, 90 141-142
temperature and pressure effects, 381 ASTM D 4045, 99 Carbon-to-hydrogen weight ratio, 11
Association parameter, 347 ASTM D 4052, 93, 144 definition, 36
ASTM, definitions of terms, 397-400 ASTM D 4124, 96 Carnahan-Starling equation of state,
ASTM D 56, 133 ASTM D 4530, 141 214-215
ASTM D 86-90, 92, 100-106, 108, 110, ASTM D 4737, 144 Cavett method, 61
113-115, 118, 131, 134, 140, 144, ASTM D 4953, 131 Cementation factor, 351
313-314 ASTM D 5296, 94 Cetane number, petroleum fractions,
ASTM D 88, 23 ASTM D 5985, 135 137-138

401
402 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

CH weight ratio, pure hydrocarbons, single carbon number groups, E

prediction, 68-69 characteristics, 161-163
Chapman-Enskog equation, 346 sulfur content estimation, 191-192 Edmister method, 65
Chapman-Enskog theory, 339 vapor pressure, 313-315 Elemental analysis, petroleum ractions,
Characterization method, evaluation viscosity, 338 98-99
criteria, 75-76 Cryoscopy, 94 EN 238, 10
Chemical potential, mixtures, 254-255 C6+ fraction,subitem refractive index, 180 End point, 88
Chen correlation, 323 C7+ fraction Energy, units, 22
Chen-Othmer correlation, 347 carbon number range approach, 186 Enthalpy, 12, 315-318
Chromatography, 96-98 comparison of distribution models, calculation diagram, 318
Chueh-Prausnitz relation, 210 179-180 ideal gas, constants, 246-247
Chung's model, 386 probability density functions, 370 two petroleum fractions, 316-317
Clapeyron equation, 252, 307-309 C8 hydrocarbons, properties, 48 Enthalpy departure, 317
Clausius-Clapeyron equation, 252 Cubic equations of state, 204-210, 319 Enthalpy of vaporization, 322
Cloud point application to mixtures, 209-210 versus temperature, 323
calculation, 382-385 other types, 208-209 Entropy, 234
petroleum fractions, 135-136 Peng-Robinson equation, 205-206, 208 ideal gas, constants, 246-247
Coal liquid fractions, heat of vaporization, Redlich and Kwong equation, 205, Entropy departure, 237
324 226-227 hard-sphere fluids, 286-287
Coefficient of thermal expansion, 236 Soave modification of Redlich and Entropy of vaporization, 252
Colloidal model, 375-376 Kwong equation, 205,208 Equations of state, 199-204
Composition, units, 21-22 solution, 206-207 corresponding state correlations,
Compressibility factor, 203, 215-221,289 unified form, 206 215-221
Consistency test, predicted physical van der Waal equation, 204-205 fugacity coefficient calculation,
properties, 71, 73 volume translation, 207-208 255-256
Continuous mixture characterization Cycloalkanes, 4 ideal gas law, 203
approach, petroleum fractions, intermolecular forces, 202-203
187-189 real gasses, 203-204
Correlation index, 122-124 D refractive index based, 225-227
Corresponding states principle, 215 velocity of sound based, 286-287
COSTALD correlation, 224 Daubert's method, 103-106 see also Cubic equations of state;
Cracking, 7 Deasphalted oils, 378 Noncubic equations of state
Cricondentherrn temperature, 202 Decane, equilibrium ratios, 282 Equilibrium flash vaporization, petroleum
Critical compressibility factor Defined fraction, 114 fractions, 91-92
definition, 32 Defined mixtures, 114-115 Equilibrium ratios, 12, 14, 269-276
prediction, pure hydrocarbons, 63-64 Definition of basic properties, 31 Ethane
Critical constants, definition, 32-33 Degrees of freedom, 199 compressibility factor, 289
Critical density, definition, 32 Density, 11,300-305 equilibrium ratios, 272
Critical point, 200 definition, 31 saturation curves, 209
Critical pressure, 11 gases, 300 Ethers, octane number, 139
n-alkyl, 52 liquid petroleum fractions, 223-224 Excess property, 249
definition, 32 liquids, 300-304 Exponential model, hydrocarbon-plus
estimation, 78-80 petroleum fractions, 93 fractions, 165-167
heavy hydrocarbons, 52-53 pure hydrocarbons, prediction, 66 Extensive property, 198-199
PNA hydrocarbons, 52 solids, 304-305 Eyring rate theory, 347
prediction, pure hydrocarbons, 60-62 units, 20-21
Critical properties Dew point, 201-202
coal liquids, 62 calculations, 371-372 F
estimation, 115-116 Diesel fuel, characteristics, 143
internal consistency, 51 Diffusion coefficients, 12, 345-351 Fenske Equation, 14
Critical temperature, 11 measurement in reservoir fluids, 354-356 Flame ionization detector, 90
n-alkanes, 50 multicomponent systems, 350 Flammability range, definition, 34
definition, 32 order of magnitude, 346 Flash calculations, 368-370
estimation, 78-80 porous media, 350-351 Flash point, 11
heavy hydrocarbons, 52-53 units, 23-24 definition, 34
influence, 13-14 Diffusivity, 12 petroleum fractions, 133-135
prediction, pure hydrocarbons, relation to refractive index parameter, Fluid properties, use of sound velocity,
60-62 353 284-292
Critical viscosity, 334 Dipole forces, 45 Fluidity, relation to refractive index
Critical volume, 11 Dipole moments, 375 parameter, 352
estimation, 79 Distillation, simulated, by gas Fluids
prediction, pure hydrocarbons, 62-63 chromatography, petroleum Newtonian and non-Newtonian, 331
Crude oils, 5-7 fractions, 89-91 wettahility, 358
API gravity, 156 Distillation curves, 11 Force, units, 19
asphaltene content, 374-378, 387-388 interconversion, 101-108 Fractured reservoirs, idealized, 391
assays, 154, 156-159 at reduced pressures, 106-108 Free-volume theory, 347
cloud point temperature, 383-384 summary chart, 109 Freezing point, 259-260
composition and properties, 6-7 petroleum fractions, 88-93 definition, 34
from atmospheric separator, 7 prediction, 108-111 depression, 281-283
lumping scheme, 186 at reduced pressures, petroleum petroleum fractions, 136-137
nomenclature, 152-153 fractions, 92-93 prediction, pure hydrocarbons, 68-70
products and composition, 9 sub- or superatmospheric pressures, 108 saturated liquid and solid properties,
properties calculation, 189-191 Double-bond equivalent, 45 304
resin content, 374-375, 387-388 Dry gas, 6 temperature, 200
 INDEX 403

Fugacity, 187-188, 237-238, 253, H Hydrogen, prediction in petroleum

382-383 fractions, 127
asphaltene, 386 Hall-Yarborough method, 63 Hydrogen sulfide, equilibrium ratios, 283
calculation from Lewis rule, 256 Hammerschmidt equation, modified, 390
coefficient, mixtures, 254-255 Hard-sphere fluids, entropy departure,
liquids, 268 286-287
mixtures, 254-255 Hard-sphere potential, 202 Ideal gas
pure gases and liquids, 256-257, 268 Heat capacity, 12, 235 mixture, heat capacity, 244
of solids, 261-263 estimation from refractive index, 321-322 thermodynamic properties, 241-247
Fugacity coefficients, 12, 238 ideal gas, constants, 246-247 Ideal gas law, 203, 209
calculation from equations of state, mixture, 250 In-situ alteration, 2
255-256 thermodynamic properties, 319-321 Infrared spectroscopy, 97
Fusion curve, 200 Heat capacity coefficients, 320 Intensive property, 198-199
Fusion line, 251 Heat capacity ratio, 235 Interracial tension, s e e Surface/interracial
Heat of combustion, 12, 324-326 tension
Heat of formation, 12 Intermolecular forces, 43, 202-203
G Heat of fusion, 201,259-261 Internal energy, 199
Heat of mixing, 249 IP 2/98, 144
Gamma density function, molar Heat of reaction, 12 IP 12, 144
distribution, 168-169 Heat of sublimation, 314 IP 13/94, 144
Gamma distribution model, 167-170 Heat of vaporization, 12, 201,252, 321-324 IP 14/94, 144
Gas chromatography, 96-97 at boiling point, 323 IP 15, 135, 144
simulated distillation, petroleum Heating value, 25, 324-326 IP 16, 136, 144
fractions, 89-91 Heats of phase changes, 321-324 IP 34/97, 144
Gas condensate system Heavy hydrocarbons IP 57, 142
C7+ fraction characteristics, 171 API gravity and viscosity, 59-60 IP 61, 99
pseudocritical properties, 160-161 atmospheric critical pressure, 51 IP 69/94, 144
SCN group prediction, 166-167 boiling point, 50, 52 IP 71/97, 144
Gas constant, 22, 24 constants, 50-51, 54 IP 107, 99, 144
Gas injection projects, 390-391 critical pressure, 52-53 IP 123/99, 144
Gas mixtures critical temperature, 52-53 IP 156/95, 144
properties, 120 prediction of properties, 50-54 IP 196/97, 144
viscosity, 335 refractive index and viscosity, 44 IP 218, 138
Gas phase, 200 Heavy petroleum fractions IP 219, 135, 144
Gas solubility, in liquids, 266-269 enthalpy, 316 IP 236, 144
s e e a l s o Vapor-liquid equilibria molecular weight and composition, 116 IP 365, 93, 144
Gas-to-liquid ratio, 337-338 Helmhohz free energy, 235 IP 370/85, 144
Gas-to-oil ratio, 368-370 Henry's constant, 267, 269 IP 380/98, 144
units, 24 Henry's law, 266-269 IP 402, 131
Gases Heptane, equilibrium ratios, 279 IP 406/99, 144
density, 300 Hexane ISO 2049, 144
diffusivity equilibrium ratios, 278 ISO 2185, 144
at high pressures, 348-350 vapor pressure, 311 ISO 2192, 144
low pressures, 346-347 n-Hexatriacontane ISO 2592 34
thermal conductivity, 339-342 acentric factor, 65 ISO 2719 144
Gasoline, characteristics, 143 critical properties, 64 ISO 2909 123
Gaussian quadrature approach, splitting, High performance liquid chromatography, ISO 2977 144
185-186 97 ISO 3007 144
Gel permeation chromatography, 94 High-shrinkage crude oil, 6 ISO 3013 144
Generalized correlation, 215 Hoffman correlation, 271-272 ISO 3014 142
Generalized distribution model, Hydrate inhibitors, 389-390 ISO 3015 135, 144
170-184 Hydrates, formation, 388-390 ISO 3016 135, 144
boiling point, 178 Hydrocarbon-plus fractions, 153, 164-184 ISO 3104 100, 144
calculation of average properties, boiling point and specific gravity ISO 3405 144
175-177 prediction, 173 ISO 3837 144
subfractions, 177-178 calculation of average properties, ISO 4262 144
C6+ fraction, 180 175-177 ISO 4264, 144
C7+ fractions, 179-180 exponential model, 165-167 ISO 5163, 144
model evaluations, 178-180 gamma distribution model, 167-170 ISO 6615, 144
prediction using bulk properties, general characteristics, 164-165 ISO 6616, 144
181-184 generalized distribution model, 170-184 ISO 6743/0, 10
probability density function, molar distribution, 167, 172-173 ISO 8708, 144
174-175 molecular weight variation, 165 ISO 12185, 93
specific gravity, 179 prediction of PDF, 173-174 Isofugacity equations, 383
versatile correlation, 170-174 probability density functions, 164-165 Isoparaffins, 3
Gibbs energy, 263 subfractions, calculation of average Isothermal compressibility, 236
binary system, 263-264 properties, 177-178
excess, 257-258 Hydrocarbons, 3-5
Gibbs free energy, 12, 235 groups, 3
Gilliland method, 347 liquid specific gravity, temperature
Glaso's correlation, 338 effect, 301 Jenkins-Walsh method, 128-129
Glossary, ASTM definitions, 397-400 pure, s e e Pure hydrocarbons Jet fuel
Goossens method, 127-128 research octane number, 140 characteristics, 143
Grouping, 184 Hydrodynamic theory, 347 enthalpy, 318
404 C H A R A C T E R I Z A T I O N A N D P R O P E R T I E S OF P E T R O L E U M F R A C T I O N S

Jossi's correlation, 337 Mixtures O

Joule-Thomson coefficient, 236 phase equilibria, 254-263
activity coefficients, 254-255, Octane
257-261 equilibrium ratios, 280
K criteria, 263-265 liquid heat capacity, 291
fugacity and fugacity coefficients, Octane number
Kay's mixing rule, 220, 372 254-257 definition, 34-35
Kesler-Lee method, 79, 81 fugacity of solids, 261-263 petroleum fractions, 138-141
Kinematic viscosity, 331,337 property change due to mixing, 249-251 Oil, speed of sound in, 286
definition, 33-34 thermodynamic properties, 247-251 Oil field, 2
estimation, 118-119 Molar density, units, 20-21 Oil reserves, 2
prediction, pure hydrocarbons, 70-73 Molar distribution, gamma density Oil wells
units, 23 function, 168-169 history, 2
Korsten method, 65, 81 Molar refraction, 47, 225 number of, 3
Kreglweski-Kay correlation, 372 Molar volume, 259-260 Oils, enhanced recovery, 390-391
Kuwait crude oil, characterization, 190 units, 20 Olefins, 4
Molecular types, characterization Oleum, 1
parameters, 121-124
L Molecular weight, 11
comparison of distribution models, 178 P
Lee-Kesler correlation, 239 definition, 31
Lee-Kesler method, 56, 60-61, 64-65, estimation, 115-116 P-T diagrams, 372-373
80-81 evaluation of methods, 76-77 Packing fraction, 214
Length, units, 18 petroleum fractions, 93-94 Parachor, 358-359
Lennard-Jones model, 202 prediction, pure hydrocarbons, Paraffins, 3-4
Lennard-Jones parameters, velocity of 55-58 content and research octane number,
sound data, 288-289 units, 19 141
Lewis rule, fugacity calculation, 256 Moles, units, 19 properties, 48
Linden method, 137 Motor octane number, 34-35, 138 Partial molar properties, mixtures, 248-249
Liquid chromatography, 90, 97 Multicomponent systems, diffusion Partial specific property, 248
Liquid density coefficients, 350 Pedersen exponential distribution model,
effect of pressure, 223-225 Multisolid-phase model, 378, 167
pressure effect, 302 382-385 Peng-Robinson equation of state, 205-206,
temperature effect, 303 208
Liquid mixtures, properties, 11%120 velocity of sound data, 289-292
Liquids N Pentane, equilibrium ratios, 276-277
density, 300-304 Percent average absolute deviation, 75
diffusivity n-d-M method, 126-127 Petroleum, formation theories, 2
at high pressures, 348-350 Naphthalene, solubility, 277-278 Petroleum blends, volume, 251
at low pressure, 347-348 Naphthas, 9 Petroleum cuts, 8
fugacity, 268 GC chromatograph, 91 Petroleum fluids
calculation, 256-257 research octane number, 140 nature of, 1-3
gas solubility, 266-269 Naphthenes, 4 characterization, importance, 12-15
heat capacity values, 319 Natural gas Petroleum fractions, 7-10, 87-146
thermal conductivity, 342-345 hydrate formation, 388 acentric factor, estimation, 115-116
viscosity, 335-338 pseudocritical properties, 160-161 aniline point, 137
see also Vapor-liquid equilibria sulfur in, 5 average boiling point, 100-101
London forces, 45 wet and dry, 6 boiling point and composition, 121
Lubricants, 9-10 Near-critical oils, 6 boiling point and distillationcurves, 88-93
Lumping scheme, 184 Newton-Raphson method, 380 bulk parameters, 114
petroleum fractions, 186-187 Newton's law of viscosity, 331 carbon and hydrogen prediction, 127
NF M 07-048, 136 carbon number range approach, 186
NF T 60-162, 10 carbon residue, 141-142
M NF T 60-101, 93 cetane number, 137-138
Nitrogen, prediction in petroleum cloud point, 135-136
Margule equation, 261 fractions, 129-130 composition, 11
Mass, units, 18 Nomenclature, 1 compositional analysis, 95-99
Mass flow rates, units, 20 Nonane, equilibrium ratios, 281 continuous mixture characterization
Mass spectrometry, 98 Noncubic equations of state, 210-215 approach, 187-189
Maturation, 2 Carnahan-Starling equation of state, critical properties, estimation, 115-116
Maxwell's equations, 235 214-215 defined mixtures, 114-115
Melting point, 11 modified Benedict-Webb-Rubinequation density
definition, 34 of state, 214, 217-220 estimation, 117
prediction, pure hydrocarbons, 68-70 SAFT, 215 specific gravity, and API gravity, 93
pressure effect, 253-254 second virial coefficients, 211-212 diesel index, 137-138
Metals, in petroleum fractions, 99 truncated virial, 212-213 distillation
Methane virial equation of state, 210-214 at reduced pressures, 92-93
compressibility factor, 289 Nonfuel petroleum products, 9-10 columns, 8
equilibrium ratios, 271 Nonhydrocarbon systems, extension of curve prediction, 108-111
hydrate formation, 388 correlations, 54-55 elemental analysis, 98-99
P-H diagram, 263-264 Nonpolar molecules, potential energy, elemental composition prediction,
speed of sound in, 286 45-46 127-130
Micellar model, 375-376 Nonwetting fluid, 357 equilibrium flash vaporization, 91-92
Miller equation, 306 Numerical constants, 24 flash point, 133-135

freezing point, 136-137
gas mixtures, properties, 120
Gaussian quadrature approach, 185-186
ideal gas properties, 243-244
interconversion of distillation data,
101-108
kinematic viscosity, estimation, 118-119
laboratory data analysis, 145-146
liquid mixtures, properties, 119-120
lumping scheme, 186-187
matrix of pseudocomponents, 111-112
method of pseudocomponent, 114-115
minimum laboratory data, 143-145
molecular type prediction, 121-124
molecular weight, 93-94
molecular weight estimation, 76
narrow versus wide boiling range
fractions, 1t2-114
nomenclature, 87
octane number, 138-141
olefin-free, 115
PNA analysis, 98
PNA composition, prediction, 120-127
pour point, 135-136
predictive method development, 145-146
pseudocritical properties, estimation,
115-116
Rackett equation, 223
refractive index, 94-95
estimation, 117
Reid vapor pressure, 131-133
simulated distillation by gas
chromatography, 89-91
smoke point, 142
specific gravity, estimation, 117
splitting scheme, 184-186
sulfur and nitrogen prediction, 129-130
surface/interracial tension, 359-360
thermodynamic properties, general
approach, 298-300
true boiling point, 89
types of composition, 96
undefined mixtures, 114
vapor pressure, 312-314
viscosity, 99-100
using refractive index, 338
V/L ratio and volatility index, 133
Winn nomogram, 74
Petroleum processing, 17
Petroleum production, 17
Petroleum products
nonfuel, 9-10
quality, 143
vapor pressure, 313-314
Petroleum waxes, 10
Phase equilibrium, 365-393
asphaltene, precipitation, solid-liquid
equilibrium, 385-388
enhanced oil recovery, 390-391
mixtures, 254-263
activity coefficients, 254-255, 257-261
criteria, 263-265
fugacity and fugacity coefficient,
254-257
fugacity of solids, 261-263
nomenclature, 365-366
pure components, 251-254
types of calculations, 366-367
vapor-solid equilibrium, 388-390
viscosity, 367-373
see also Vapor-liquid-solid
equilibrium-solidprecipitation
Phase rule, 199
Physical properties, 10-12
Planck constant, 24
PNA analysis, 98
PNA composition, prediction, 120-127
PNA three-pseudocomponent model, 115
Polarizability, 47
Porous media, diffusion coefficients,
350-351
Potential energy, nonpolar molecules,
45-46
Potential energy function, 202
Potential energy relation, two-parameter,
46, 48
Pour point, 11
petroleum fractions, 135-136
Poynting correction, 257
Prandtl number, 339
Pressure
triple point, 199
units, 19
Propane
compressibility factor, 289
equilibrium ratios, 273
The Properties of Gases and Liquids, 16
Properties of Oils and Natural Gases, 16
Pseudocomponent method, 320
Pseudocomponent technique, 112
Pseudocomponents
generation from Gaussian quadrature
method, 185-186
matrix, 11 I-112
Pseudocritical properties, 12, 32
gas condensate, 160-161
natural gas, 160-161
Pseudoization, 184
Psuedocomponents, 13
Pure components, vapor pressure,
305-306
Pure compounds
critical thermal conductivity, 241
liquid thermal conductivity, 343
vapor pressure, coefficients, 308-309
viscosity coefficients, 333-334
Pure gases, fugacity, 268
calculation, 256-257
Pure hydrocarbons, 30-83
acentric factor, prediction, 64-65, 81
boiling point, prediction, 58-59
CH weight ratio, prediction, 68-69
characterization, 45-55
parameters, 48-50
criteria for evaluation of characterization
method, 75-76
critical temperature and pressure,
prediction, 60-62
critical volume, prediction, 62-63
data sources, 36-37
definition of properties, 31-36
density, prediction, 66
estimation of critical properties, 77-81
extension of correlations to
nonhydrocarbon systems, 54-55
freezing/melting point, prediction, 68-70
generalized correlation for properties,
45-48
heavy, properties, 37, 44-45
kinematic viscosity, prediction, pure
hydrocarbons, 70-73
molecular weight prediction, 55-58
nomenclature, 30
prediction of properties, recommended
methods, 83
properties, 37-43
refractive index, prediction, 66-68
secondary properties, 41-43
INDEX 4 0 5
specific gravity/API gravity
prediction, 58-60
Winn nomogram, 73-75
see also Heavy Hydrocarbons
PVT relations, 199-202
critical point, 46
intermolecular forces, 202-203
nomenclature, 197-198
Rackett equation, 222-225
Q
Quadratic mixing rule, 209
R
Rachford-Rice method, 368
Rackett equation, 222-225, 301
pressure effect on liquid density, 223-225
pure component saturated liquids,
222-223
Rackett parameter, 222
Raoult's law, 188, 265-267
Real gases, equations of state, 203-204
Redlich-Kister expansion, 257
Redlich-Kwong equation of state, 46, 205,
226-227, 300
velocity of sound data, 289-292
Refining processes, 7
Refractive index, 11
basis for equations of state, 225-227
C6+ fraction, 180
definition, 32
estimation, 117
heat capacity estimation from, 321-322
heavy hydrocarbons, 44
parameter
relation to fluidity, 352
relation to diffusivity, 353
petroleum fractions, 94-95
pure hydrocarbons, prediction, 66-68
Refractivity intercept, 11
definition, 35
Reid vapor pressure, 11, 33, 131-133
Reidel method, 63
Relative volatility, 14
effect of error, 14
Research octane number, 34-35, 138
Reservoir fluids, 2, 5-7
composition and properties, 6-7
C7+ fractions, characteristics, 163-164
definition, 5
diffusion coefficients measurement,
354-356
flash calculation, 369
laboratory data, 153-155
lumping scheme, 186
nomenclature, 152-153
properties calculation, 189-191
single carbon number groups,
characteristics, 161-163
types and characteristics, 6
Residual enthalpy, 237
Residual Gibbs energy, 237-238
Residual heat capacity, 238
Resins, 374-375
Retention time, 90
Retrograde condensation, 202
Riazi-Daubert correlations, 58, 78-80
Riazi-Daubert methods, 55-57, 58-60, 62,
102-103, 124-126
Riazi-Faghri method, 341,343
406 CHARACTERIZATION AND PROPERTIES OF PETROLEUM FRACTIONS
Riazi method, 127
Riedel equation, 313,323
Riedel method, 342
S units, 24
SAFT model, 386
Saturation curves, ethane, 209
Saturation pressure, thermodynamic
properties, 251- 254
Saybolt viscosity, 35
Scatchard-Hildebrand relation, 258, 261
Schmidt number, 345
SCN groups
characteristics, 161
exponential model, 165-167
molecular weight boundaries, 168
Self-diffusion coefficient, 345
Sensitivity of fuel, 138
Separation by solvents, 96
Shear stress, 331
Shift parameter, 208
SI units, 18
Size exclusion chromatography, 93-94
Smoke point, petroleum fractions, 142
Solid-liquid equilibrium, 385-388
Solid solubility, 276-281
Solid solution model, 378, 380-382
Solids
density, 304-305
fugacity calculation, 261-263
vapor pressure, 314-316
Solubility, 259-260
Solubility parameter, units, 24
Solvents, 9
Soreide correlation, 58
Sound velocity
equations of state based on, 286-287
Lennard-Jones and van der Waals
parameters, 288-289
prediction of fluid properties,
284-292
RK and PR EOS parameters, 289-292
virial coefficients, 287-288
Specific energy, units, 22
Specific gravity, 11
comparison of distribution models,
178-179
definition, 31
estimation, 117
hydrocarbon-plus fractions, 173
hydrocarbons, temperature effect, 301
petroleum fractions, 93
prediction, pure hydrocarbons, 58-60
units, 21
Specific volume, units, 20
Spectrometric methods, 98
Speed of light in vacuum, 24
Splitting scheme, petroleum fractions,
184-186
Square-Well potential, 202
Standing-Katz chart, 215-216
Stiel-Thodos method, 341
Stokes-Einstein equation, 349
Sublimation, 314
Sublimation curve, 200
Sublimation line, 251
Sublimation pressure, 315
Sulfur
crude oil content, 191-192
in natural gas, 5
prediction in petroleum fractions,
129-130
Supercritical fluid, 200 U
Surface/interfacial tension, 12,
356-361 Units
predictive methods, 358-361 composition, 21-22
theory and definition, 356-358 conversion, 25
density, 20-21
diffusion coefficients, 23-24
energy, 22
force, 19
fundamental, 18
Temperature gas-to-oil ratio, 24
triple point, 199 importance and types, 17-18
units, 19, 19-20 kinematic viscosity, 23
Tensiometer, 357 length, 18
Thermal conductivity, 12, 339-345 mass, 18
critical, 341 mass flow rates, 20
gases, 339-342 molar density, 20--21
liquids, 342-345 molecular weight, 19
versus temperature, 340 moles, 19
units, 23 prefixes, 18
Thermal conductivity detector, 90 pressure, 19
Thermodynamic properties, 232-292 rates and amounts of oil and gas,
boiling point, elevation, 282-284 24-25
calculation for real mixtures, 263 solubility parameter, 24
density, 300-305 specific energy, 22
departure functions, 236-237 specific gravity, 21
enthalpy, 31 6-318 surface tension, 24
freezing point depression, 281-283 temperature, 19-20
fugacity, 237-238 thermal conductivity, 23
generalized correlations, 238-241 time, 18-19
heat capacity, 319-321 viscosity, 23
heat of combustion, 324-326 volume, 20
heat of vaporization, 321-324 volumetric flow rates, 20
heats of phase changes, 321-324 UOP characterization factor, 13
ideal gases, 241-247
measurable, 235-236
mixtures, 247-251 V
nomenclature, 232-234
property estimation, 238 Van der Waal equation, 204-205
residual properties, 236-237 Van der Waals parameters, velocity of
saturation pressure, 251-254 sound data, 289
solid-liquid equilibria, 276-281 Van Laar model, 257-258
summary of recommended methods, Vapor, 200
326 Vapor-liquid equilibria, 25t-253, 265-276
use of sound velocity, 284-292 equilibrium ratios, 269-276
vapor-liquid equilibria, 265-276 formation of relations, 265-266
Thermodynamic property, 199 Raoult's law, 265-266
Time, units, 18-19 solubility of gases in liquids, 266-269
Toluene, effect on asphaltene precipitation, Vapor-liquid equilibrium calculations,
377-378 367-373
Tortuosity, 350-351 bubble and dew point calculations,
Transport properties, 329-362 370-372
diffusion coefficients, 345-351 gas-to-oil ratio, 368-370
diffusivity at low pressures P-T Diagrams, 372-373
gases, 346-347 Vapor liquid ratio, volatility index and, 133
liquids, 347-348 Vapor-liquid-solid equilibrium-solid
diffusivity of gases and liquids at high precipitation, 373-385
pressures, 348-350 heavy compounds, 373-378
interrelationship, 351-354 wax precipitation
measurement of diffusion coefficients in multisolid-phase model, 382-385
reservoir fluids, 354-356 solid solution model, 378, 380-382
nomenclature, 329-330 Vapor pressure, 11,200, 305-316
surface/interfacial tension, 356-361 Antoine coefficients, 310
thermal conductivity, 339-345 definition, 33
viscosity, 331-338 petroleum fractions, 312-314
Triple point pressure, 199 predictive methods, 306-312
Triple point temperature, 199 pure components, 305-306
Trouton's rule, 322 pure compounds, coefficients, 308-309
True boiling point, distillation curve, 182 solids, 314-316
True critical properties, 372-373 Vapor pressure method, 94
Tsonopoulos correlations, 62 Vapor-solid equilibrium, 388-390
Two petroleum fractions, enthalpy, Vignes method, 347
316-317 Virial coefficients, velocity of sound data,
Twu method, 61-62, 80 287-288
 INDEX 407

Virial equation of state, 210-214 Volatility index, and vapor liquid ratio, 133 Wax precipitation
truncated, 240 Volume, units, 20 rnultisolid-phase model, 382-385
Viscosity, 12, 331-338 Volume translation, cubic equations solid solution model, 378,
gases, 331-335 of state, 207-208 380-382
heavy hydrocarbons, 44 Volumetric flow rates, units, 20 Waxes, 373
liquids, 335-338 Wet gas, 6
petroleum fractions, 99-100 Wetting liquid, 357
pressure effect, 334 W Wilke-Chang method, 347
versus temperature, 332 Wilson correlation, 273
units, 23 Walsh-Mortimer method, 137 Winn method, 137
Viscosity-blendingindex, 337 Water Winn-Mobil method, 62
Viscosity coefficients, pure liquid ideal gas heat capacity, 242-243 Winn nomogram, 73-75
compounds, 336-337 vapor pressure, 312 Won model, 380
Viscosity gravity constant, 11 Watson characterization factor, 320
definition, 35-36 Watson K, 11, 13, 323
Viscosity index, 122-124 definition, 35 X
Viscosity-temperature relation, 72 Wax appearance temperature, 378,
Volatility, properties related to, 131-135 382 Xylene, vapor pressure, 311